STUDIES CONCERNING STEREOSELECTIVE,
REGIOSELECTIVE AND CATALYTIC ORGANOMETALLIC
REAGENTS IN ORGANIC SYNTHESIS
A. THESIS
Presented to
The Faculty of the Division of Graduate
Studies and Research
by
Stephen A. Noding
In Partial Fulfillment
of the Requirements for the Degree
Doctor of Philosophy
in the School of Chemistry
Georgia institute of Technology
Decembe.r 1978
James -A. Stanrielo f
r v - Dr. Er ng GroveTitellt;''Ji4:
ate Appr"oven by Chairman
STUDIES CONCERNING STEREOSELECTIVE,
REGIOSELECTIVE AND CATALYTIC ORGANOMETALLIC
REAGENTS IN ORGANIC SYNTHESIS
Approved:
Dr. Etigetfie C. Ashby, 'AdArisor
ACKNOWLEDGMENTS
The author wishes to express his sincere appreciation to his
advisor, Dr. Eugene C. Ashby, for his suggestion of these problems and
for his guidance, patience and continuing encouragement throughout the
course of this study. The author also wishes to thank the other members
of the Reading Committee, Dr. Erling Grovenstein, Jr., Dr. Herbert 0.
House and Dr. James A. Stanfield, for their helpful comments during the
preparation of this thesis. For their suggestions during many discussions
concerning the work in this thesis, the author wishes to thank his co-
workers, especially R. Scott Smith, Dr. Tim Smith and Dr. J. J. Lin.
The author has held Union Camp Corporation and Alcoa fellowships
for which he is grateful. Financial assistance by the Georgia Institute
of Technology and the National Science Foundation is also gratefully
acknowledged.
The author would like to dedicate this thesis to his mother, Vera
E. Noding Langfeldt, and in loving memory to his father, Alfred L. Noding,
who provided him with the opportunity and incentive to attend high school
and college. Any success the author has or will have is based upon the
inspiration and foundation they laid.
ii
TABLE OF CONTENTS
Page
• ACKNOWLEDGMENTS • ... • • • • • • . ..... .. OOOOO . • • • •
LISTOF TABLES OOOOOO . • . • .•. . • • • .. . OOOO • . viii •
LIST OF ILLUSTRATIONS xiii
SUMMARY xv
PART I
A STUDY OF STERIC APPROACH CONTROL
VERSUS
PRODUCT DEVELOPMENT CONTROL
Chapter
. INTRODUCTION 2
Background , Purpose
II. EXPERIMENTAL 5
General Considerations Materials Apparatus Analytical Preparations
III. RESULTS AND DISCUSSION 32
Synthesis of Model Systems for Reduction Studies Stereochemistry of 7-Norbornanone Reduction Synthesis of Model Systems for Alkylation Studies Stereochemistry of 7--Norbornanone Alkylation
IV. CONCLUSTIONS ....... 0 • • • •• OOOOOO _• • • • - 48
iii
LITERATURE CITED 56
TABLE OF CONTENTS
Page
PART II
HYDROMETALLATION OF ALKYNES AND ALKYNES
CATALYZED BY TRANSITION METAL HALIDES
Chapter
. INTRODUCTION 59
Background Purpose
II. EXPERIMENTAL 61
Apparatus Analytical Materials General Reactions of Alkenes and Alkynes General Quenching Techniques General Reactions of Complex Metal Hydrides General Reactions of LiH and NaH General Reactions for the Carbonylation of Simple and Mixed
Metal Hydrides
III. RESULTS AND DISCUSSION 79
Reactions of Alkenes Reactions of Dienes Reactions of Alkynes Survey of Catalysts for Hydrometallation of Internal Alkenes Further Reactions with Carbonyl Compounds and Oxygen Survey of Substituted Alanes Hydrometallation with LiH Main Group Complex Metal Hydride Reactions Reductions Using HCo(C0) 4 Simple and Complex Metal Hydride Carbonylations
IV. CONCLUSIONS 116
LITERATURE CITED 153
TABLE OF CONTENTS
Page
PART III
REACTIONS OF MAGNESIUM HYDRIDE:
STEREOSELECTIVE REDUCTION OF CYCLIC AND
BICYCLIC KETONES BY LITHIUM ALKOXYMAGNESIUM HYDRIDES
Chapter
I. INTRODUCTION. . . OOOOOOOOOOOOOOO • . • . 158
Background Purpose
II. EXPERIMENTAL 160
General Considerations Analyses Materials General Reaction of Hydrides with Ketones Qualitative Rate Studies
III. RESULTS AND DISCUSSION 168
IV. CONCLUSION 179
LITERATURE CITED 196
TABLE OF CONTENTS
Page
PART IV CONCERNING SALT EFFECTS:ON THE STEREOSELECTIVITY OF
ORGAMMETALLIC COMPOUND ADDITION TO KETONES
Chapter
I. INTRODUCTION 199
Background Purpose
II. EXPERIMENTAL 203
Apparatus Analytical Materials General Reactions of Ketones
III. RESULTS AND DISCUSSION . 209
IV. CONCLUSIONS 218
LITERATURE CITED 230
TABLE OF CONTENTS
Page
PART V
ALKYLATIONS OF ENONES AND. KETONES USING
SUBSTITUTED ALKYLAIMIINUM COMPOUNDS
Chapter
I. INTRODUCTION 233
Background Purpose
II. EXPERIMENTAL 235
General Considerations Analytical Materials General Reactions of Enones General Reactions of Ketones
III. RESULTS AND DISCUSSION 243
IV. CONCLUSION. . . . . ........ . . . . ..... 255
LITERATURE CITED 276
vii
LIST OF TABLES
PART I
A STUDY OF STERIC APPROACH CONTROL
VERSUS
PRODUCT DEVELOPMENT CONTROL
Table Page
1. Reactions of LiA1H 4 with Ketones I, II and III in Diethyl Ether and THE 49
2. Reactions of Group IIIb Metal Hydrides with Ketones (II) and (III) in THE 50
3. Reactions of Common Cation Complex Metal Hydride with Ketones (II) and (III) in THE 51
4. Reactions of Varying Cations of Complex Metal Hydrides with Ketones (II) and (III) in THE. . . . . . . . ........ 52
5. Methylmagnesium Bromide Reactions with Ketones (I), (II) and (III) in Diethyl Ether and THE 53
6. Reactions of Alkylmetal Reagents with Ketones (II) and (III) in Diethyl Ether 54
7. Reactions of RigX Compounds with Exo-2-Methyl-7-Norbornanone (II) in E0 Solvent at Room Temperature for 30 Hours in 2:1 Molar Ratio 55
PART II
HYDROMETALLATION OF ALKENES AND ALKYNES
CATALYZED BY TRANSITION METAL HALIDES
8. Reactions of 1-Octene with HA1(NPt) 2 in the Presence of 5 Mole Percent Catalyst and Quenched with D
20
viii
119
LIST OF TABLES
Table
9. Reactions of 1-Octene with HAl(NPr 1 ),.) in the Presence of 5 Mole Percent Cp2TiC12 and Quenched with D 20 120
10. Reactions of Alkenes with HAl(NPr 2)2 and 5 Mole Percent of
Cp 2TiC1 2 and Quenched with D 20 121
11. Regioselectivity in the Reaction of HAl(NPr,t) with Alkenes in the Presence of 5 Mole Percent Cp2TiC12 as betermined by Quenching with a Benzene Solution of Iodine 123
12. Reactions of Dienes with HAl(NPr2 i ) 9 in THF or Benzene at Room Temperature for 12 Hours in a Digne/HA1(NP4) 2 Ratio of 1:2 . 124
13. Reactions of Cis-2-Hexene and HAl(NPr) 2 with Various Catalysts 2 and Quenched with a Benzene Solution of Iodine 125
14. Reactions of Alkynes with HAl(NP4),, and 5 Mole Percent Cp 2TiC12 in an Alkyne/Alane Ratio oi 1.0:1.02 ..... . . . 126
15. Reactions of 2-Hexyne with HAl(NP4) 2 and Cp2TiC1 2 in 1.0: 1.02:0.1 Mole Ratio 129
16. Reactions of 1-Octene with HAl(NPrb 2-Cp 2TiC12 : Effect of Temperature and Catalyst Concentration 130
17. Reactions of the Hydrometallated Species with Carbonyls or Oxygen or Carbon Dioxide in Benzene at Room Temperature 24 Hours 131
18. Reactions of 1-Octene with Substituted Alanes in the Presence of 5 Mole % Cp 2TiC12 in Benzene at Room Temperature for 12 Hours 133
19. Reactions of LiH and Transition Metal Halide with 4-t-Butyl-cyclohexanone in a 1:1:1 Ratio at Room. Temperature for 24 Hours in THF 134
20. Reactions of Carbonyl Substrated with LiH:VC1 3 in THF at 45°C for 36 Hours in a Mole Ratio of 1•3 135
21. Reactions of Alkenes with LiH:VC13
in THE at 45 °C for 36 Hours in a Mole Ratio of 1:3 and Quenched with D 20 136
22. Reactions of LiH and NaH with 1-Octene in the Presence of Catalytic Amounts of Transition Metal Halides in Benzene at Room Temperature for 24 Hours and Quenched with D 20. . . . 137
ix
Page
LIST OF TABLES
Table Page
23. Reactions of Alkynes with LiH:VC1 3 in Benzene at 45°C for 36 Hours in a Mole Ratio of 1:3 and Quenched with D 20 . . . . 138
24. Reactions of Enones with LiH:VC13 in Benzene at 45 °C for 36
Hours in a Mole Ratio of 1.3 139
25. Reactions of Complex Aluminum Hydrides with Olefins and Alkynes in the Presence of 5 Mole % Cp 2TiC12 in THE for Two Hours in a Mole Ratio of 1:1 and Quenched with D 20. . . . 140
26. Reaction of Ketones with HCo(CO) 4 at Various Temperatures in Hexane and Ketone:HCo(C0) 4 of 1.2 151
27. Carbonylations of Simple and Complex Metal Hydrides in the Presence of 5 Mole Percent . Transition Metal Halides at 4000 psi, Room Temperature, in THF or Hexane for 20 Hours 152
PART III
REACTIONS OF MAGNESIUM HYDRIDE:
STEREOSELECTIVE REDUCTION OF CYCLIC AND
BICYCLIC KETONES BY LITHIUM ALKOXYMAGNESIUM HYDRIDES
28. Preparation of Lithium Alkoxymagnesium Hydrides [LiMgH 2 (0R)] by the Reaction of Magnesium Hydride with Lithium Alkoxides in a 1:1 Ratio 0 180
29. 'Reaction of 4-t-Butylcyclohexanone with LiMgH2 (0R) Compounds at Room Temperature in THF Solvent 181
30. Reactions of 3,3,5-Trimethylcyclohexanone with LiMgH 2 (0R) Compounds at Room Temperature in THF and 1:2 Molar Ratio. . . 182
31. Reactions of 2-Methylcyclohexanone with LiMgH2(OR) Compounds
at Room Temperature in THF Solvent in 1:2 Ratio 183
32. Reactions of Camphor with LiMgH 2 (0R) Compounds at Room Temperature in THE Solvent in 1:2 Molar Ratio 184
33. The Reaction of LiMgH9 (0-2,2,6,6-Tetrabenzylcyclohexyl) in THF Solvent with 4-t-tutylcyclohexanone at Various Temperatures and Reaction Times in 2:1 Molar Ratio 185
x
LIST OF TABLES
Table Page
34. Reactions of 4-t-Butylcyclohexanone with Metal Hydride's and Magnesium Alkoxides at Roan Temperature in THE Solvent and in 1:2 Molar Ratio for 24 Hours 186
35. Reactions of 2 -Methylcyclohexanone with Metal Hydrides and Magnesium Alkoxides at Room Temperature in THE Solvent, and in 1:2 Molar Ratio for 24 Hours 187
36. Reactions of 3,3,5-Trimethylcyclohexanone with Metal Hydrides and Magnesium Alkoxides at Room Temperature in THE Solvent and in 1:2 Molar Ratio for 24 Hours 188
37. Reactions of Camphor with Metal Hydrides and Magnesium Alkoxides at Room Temperature in THE Solvent and in 1:2 Molar Ratio for 24 Hours 189
PART IV
CONCERNING SALT EFFECTS ON THE STEREOSELECTIVITY OF
ORGANOMETALLIC COMPOUND ADDITION TO KETONES
38. Reactions of CH3Li—Metal Salts with 4-t-Butglcyclohexanone in Diethyl Ether Solvent for 2 Hours at -78 C in 2:1:1 Ratio. 220
39. ReaCtions of CH3Li-LiC10 with Various Ketones in Et20 Solvent for 2 Hours at -/4 8o C . 221
40. Rate of Reaction of Ketones with CH LiC104 at -78°C in Diethyl Ether Solvent 222
41. Reactions of RLi-LiC10 4 with 4-17Butylcyclohexanone in Et 20 Solvent for 2 Hours at -78 C 223
42. Reactions of RLi-LiC10 with 2-Methylcyclohexanone in Et20 Solvent for 2 Hours at 4
-78oC 224
43. Reactions of t-Butyllithiva with Ketones in the Presence and Absence olLiC10
4 at -78°C in Et 20 Solvent in 2:1:1 Ratio 225
44. Reactions of Me2Mg -Salt with 4 -t -Butylcyclohexanone in Et 20 Solvent for 2 Hours at -78 °C in 2:1:1 Ratio 226
xi
LIST OF TABLES
Table Page
45. Reactions of Me,)Mg-LiC10 with Ketones in Et 20 Solvent for 2 Hours at -78°t in 2:1:r Ratio 227
46. Reactions of Me lMg with 4-t-Butylcyclohexanobe in the Presence and Absence of tiC10 4
in Et20 Solvent at -78 C in 2:1:1 Ratio 228
47. Reactions of Me lAl-Salt with 4-t-Butylcyclohexanone in Et 20 Solvent for 12 Hours at -78 C in a 2:1:1 Ratio 229
PART V
ALKYLATIONS OF ENONES AND KETONES USING
SUBSTITUTED ALKYLALUMINUM COMPOUNDS.
48. Reactions of MenAIK3-n
Compounds with Enone (I) 256
49. Reactions of Me2A1I with Other Enones in Benzene and THE at
Room Temperature for 24 Hours in a 2:1 Ratio 260
50. Reactions of Me3Al with Enone (I) and Enone (II) in the Presence of Coordinating Agents at Room Temperature for 24
261 Hours
Et n AIK3-n Compounds with Enone (II) 262 51. Reactions of
PhnAIX
3-n Compounds with Enone (II) 265 52. Reactions of
xii
53. Reactions of Ketone (I)
MenADC
3-n Compounds with 4 -t -Butylcyclohexanone,
269
54. Reactions of Me Alwith Ketone (I) in the Presence of Co- ordinating Agents at Room Temperature for 24 Hours in a 1:1:1 Ratio 273
55. Reactions of EtnAIX3-n Compounds with 4-t-Butylcyclohexanone, Ketone (I) 274
56. Reactions of PhnAIX3-n Compounds with 4 -t -Butylcyclohexanone, Ketone (I) 275
LIST OF ILLUSTRATIONS
PART I
A STUDY OF STERIC APPROACH CONTROL
VERSUS
PRODUCT DEVELOPMENT CONTROL
Scheme
1.
2.
Preparation of 7-Norbornanone, (I)
Preparation of Exo-2.41ethyl-7-Norbornanone (II) and Endo-2-
Page
33
Methy1-7-Norbornanone (III) 34
3. Alternate Synthetic Route to Exo-2-Methyl-7-Norbornanone (II) and Endo-2-Methy1-7-Norbornanone (III) 36
4. Proposed Synthetic Scheme for the Preparation of Exo-2-Methy1- 7-Methyl-Anti-7-Norbornanol (XVIIIb) 42
5. Proposed Mechanism for the Acid Isomerization of (XVIIIa) to (XVIIIb) 43
PART II
REACTIONS OF MAGNESIUM HYDRIDE:
STEREOSELECTIVE REDUCTION OF CYCLIC AND
BICYCLIC KETONES BY LITHIUM ALKOXYMAGNESIUM HYDRIDES
6. Proposed Catalytic Hydrometallation Mechanism 80
7. Proposed Mechanism for the Catalytic Hydrometallation of 1- Phenylpropyne 91
8. Proposed Mechanism for the Polydeuterated Octane Formation. . 90
9. Proposed Mechanism for the Production of the Kinetic and Thermodynamic Products from the Catalytic Hydrometallation Reaction 102
LIST OF ILLUSTRATIONS
Scheme Page
10. Proposed Mechanism for the Reduction of Ketones with HCo(C0) 4 . 111'
PART III
REACTIONS OF MAGNESIUM HYDRIDE:
STEREOSELECTIVE REDUCTION OF CYCLIC AND
BICYCLIC KETONES BY LITHIUM ALKOXYMAGNESIUM HYDRIDES
Figure Page
1. IR Spectra of Simple and Complex Metal Alkoxides; a) Li
b) HMg
2. NMR Spectra of Simple and Complex Metal Alkoxides; a) HMg
b) LiMgH2 ( ) c) LiMgH(
3. The Reaction of LiMgH2 (0
) in THE with 4-t-Butylcyclo- -25°C; A 0°C; . RT (25°C) . . Hexanone in 2:1 Ratib: . . 194
xiv
c) LiMgH2 ( ) d) LiMgH(
) 2 d) Li
190
192:
SUMMARY
PART I. A sum OF STERIC_APPROAcH CONTROL
VERSUS
PRODUCT DEVELOPMENT CONTROL
The concept of "product development control" has been used to
explain the stereochemistry of many reactions in which the observed
isomer ratio reflects the stability of the product. This concept has
been used particularly to explain predominant formation of the most
stable isomer in reactions of LiA1H4 and MeltBr with substituted cyclo-
bexanones. A study of the reaction of LiA1H4 and MeMgEr with 7-norbor-
nanone and its exo-2-methyl and endo-2-methyl derivatives shows that
the most unstable isomer is formed exclusively and hence "product
development control" is not a factor in these reactions. In an attempt
to broaden the scope of this study, three series of reagents were
studied: (1) LiBH4 , LiA1H4 and LiGaH4 , (2) BH3 , AlH3 and GaH3 , and (3)
(CH3 ) 2Be, (CH3 ) 2Zn, (CH3 ) 2Ng and (CH3 ) 3A1. In no case was "product
development control" observed. The reactions with the 7-norbornanone
system are similar in nature to those with cyclohexanones, except that
the complicating factors of torsional strain, compression effects and
conformational changes which are present in cyclohexanone systems are
not present in the 7-norbornanone system. The concept of "product
development control" is, therefore, a questionable one in ketone reduc-
tions involving LiAIH4 and alkylations involving MeMgBr.
xv
PART II. HYDROMETALLATION OF ALKYNES AND ALKYNES
CATALYZED BY TRANSITION METAL HALIDES
We have found that bis-dialkylaminoalanes, HA1(NR 2 ) 2 , when allowed
to react with alkenes or alkynes in the presence of a catalytic amount of
Cp2TiC12 provide high yields of deuterated or iodinated hydrocarbons when
the reaction mixtures are quenched with D20 or a benzene solution of
iodine. These hydrometallated species were also allowed to react with
carbonyl compounds (e.g. benzaldehyde); however, the normal addition
products were produced in very low yields. Instead, associated tertiary
amines (PhCH2NR
2) were produced in substantial amounts. Our investiga-
tions also included other catalytic hydrometallation reactions of alkenes
and alkynes with simple and complex metal hydrides. By using Cp 2TiC12
as a catalyst with LiA1H4, NaA1H4' Vitride, LiA1(NEt 2 ) 2H2 and NaAl(NEt 2 ) 2H2'
NaAl(NEt2 ) 2H2 , quantitative yields of deuterated products were obtained
which in some cases were a distinct improvement over previous literature
reports. Stoichiometric reactions of LiH-VC13 with ketones, aldehydes
or enones were also investigated. Similar reactions with alkynes did not
take place; however, with alkenes, the reactions produced alkanes al-
though only a small amount of deuterium incorporation was observed when
the reaction was quenched with D 20.
A stereochemical study involving the reduction of cyclic ketones
with HCo(CO)4 was also undertaken. Unfortunately, the reactions were so
slow that only small amounts of the corresponding alcohols (equal ratios
of axial and equatorial alcohols) were produced.
xv i
reduced ketone (I) to pro-
xvii
A preliminary investigation of carbonylation of simple and complex
metal hydrides in the presence of transition metal halide catalysts was
also conducted. During this study no alcohol or formate products were
observed.
PART III. REACTIONS OF MAGNESIUM HYDRIDE:
STEREOSELECTIVE REDUCTION OF CYCLIC AND
BICYCLIC KETONES BY LITHIUM ALKOXYMAGNESIUM HYDRIDES
A series of lithium alkoxymagnesium hydrides, LiMgH 2 (0R), were
prepared and allowed to reduce 4-t-butylcyclohexanone (I), 3,3,5-trimethyt-
cyclohexanone (II), 2-methylcyclohexanone (III) and camphor (IV). It was
found that very bulky secondary cyclic alkoxy groups such as 2,2,6,6-tetra-
methyl- or benzylcyclohexoxy were very stereoselective in the reduction
of these ketones. For example, LiMgH2 (
vide 89% of the axial alcohol compared to HMge Qwhich provided 83%
of the axial alcohol.
The LiMgH(OR) 2 when R = or Oreagents were also found to
reduce ketones (I), (II), (III) and (IV) stereoselectively but to a lesser
extent and with more enolization than observed for the LiMgH2 (OR) reagents.
PART IV : CONCERNING SALT EFFECTS ON THE STEREOSELECTIVITY OF
ORGANOMETALLIC COMPOUND ADDITION TO KETONES
The reaction of 4-t-butylcyclohexanone with methyllithium in the
presence of LiC104 resulted in the formation of the corresponding axial
alcohol in 92% yield. This was a very unusual stereochemical observation
in that only 65% of the axial alcohol was formed in the absence of LiC104 .
This result was attributed to complexation of the ketone by LiC10 4
followed by the addition of CH3Li to the carbonyl group rather than by
addition of a CH3Li-LiC104 complex directly to the uncomplexed ketone.
To complete a more detailed investigation of this unusual result, other
salts and ketones were studied in a similar manner. In addition to
CH3Li, (CH3 ) 2Mg and (CH3 ) 3A1 were also allowed to react with 4-t-butyl-
cyclohexanone in the presence of equalmolar ratios of various salts were
studied to exam the effect on the stereochemistry of the alcohol products.
It was shown that of all the salts tested, LiC104 had the greatest
effect towards increased stereoselectivity of CH3Li reagents. The other
RLi reagents, t-butyllithium and phenyllithium, did not show as great an
effect. Also, (CH3 ) 2Mg and (CE3 ) 3A1 showed very little increased stereo-
selectivity when allowed to react with ketones in the presence of metal
salts.
PART V: ALKYLATIONS OF ENONES AND KETONES USING
SUBSTITUTED ALKYLALUMINUM COMPOUNDS
Earlier workers in this group have shown that H 2A1I provided 100%
of the 1,4-conjugate addition product when allowed to react with enones.
The possibility of using Rn AIX3-n
compounds to promote the non-catalyzed
1,4-conjugate addition to enones under consistent conditions was studied.
It was shown that for the enone reactions, the reagent which produced
the greatest amount of 1,4-coniugate addition product without the presence
of a catalyst was R2A1I; however, the yield was low.
A systematic study concerning these compounds towards the alky-
lation of model ketone systems in order to find an effective stereo-
selectivity reagent was also investigated. It was observed once again
that the most stereoselectivity reagent was R2A1I, but the reaction was
extremely slow. Therefore, it can be concluded that the use of these
reagents to effect 1,4-conjugate addition of enones or stereoselectivity
addition to ketones is impractical.
CHAPTER I
IN
Background
In recent years the area of stereoselective reduction and alkyla-
tion of ketones by metal hydrides and alkyls has received considerable
attention.1,2
All mechanisms concerning the stereoselective addition or
reduction.of ketones assume that the entering group approaches the car-
bonyl carbon's Tr* orbital on a plane perpendicular to the plane of the
carbonyl group so that maximum orbital overlap is achieved in the transi-
tion state. Dauben and co-workers3 coined the terms "steric approach
control" and product development control" and suggested that these factors
are important in determining the stereochemistry of LiA1H4 reduction of
cyclohexanones. "Steric approach control" implies an early, reactant-like
transition state in which the entering group approaches the least hindered
side of the ketone whereas "product development control" implies a late,
product-like transition state in which the observed isomer ratio reflects
the thermodynamic stability of the product.
The concept of "steric approach control" is generally agreed to be
valid since certainly the ability of one molecule to approach another
must depend to some extent on the steric requirements of the molecules
involved. However, the concept of "product development control" has
been questioned by Eliel and co-workers 4-7 on the basis of competitive
rate studies involving LiA1H 4 and 3,3,5-trimethylcyclohexanone. They
have shown that an axial methyl group in the 3 and/or 5 position retards
2
the rate of axial attack compared to 4-tert-butylcyclohexanone, whereas
the rate of equatorial attack remains essentially the sane. This ob-
servation is not consistent with that predicted by "product development
control" in that an axial methyl substituent would be expected to retard
equatorial attack.
As an alternative to "product development control", Cherest and
Felkin introduced the concept of "torsional strain"8-11
and we have
developed the concept of "compression effect" to explain the unusual
stereochemistry observed in the reactions of (CH 3 ) 3A1 with substituted
cyclohexanones.12
The cyclohexanone ring system may also undergo con-
formational changes, a factor which has been discussed by Landor and
Regan.13
Fibre recently orbital symmetry arguments 14 and unequal dis-
tortion of electron density15
about the carbonyl group have been sug-
gested to explain the stereochemistry of certain reactions.
Alkylation and reduction studies of a model ketone system in
which torsional strain, compression effects and conformation changes are
not possible were carried out so that "steric approach control" and
"product development control" could be evaluated independently of these
other possible effects. 7-Norbornanone (I) exhibits bridgehead hydrogen
atoms in the 1 and 4 positions which eclipse the carbonyl group in the
7 position. This unique feature, unlike that of the 2,6-diequatorial
hydrogens in cyclohexanone which lie 4-5° above the plane of the carbonyl
group, eliminates torsional strain or compression effect as a compli-
cating factor in evaluating stereochemical data obtained from this system.
The fact that (I) is a rigid bicyclic system further eliminates confor-
mational changes in the substrate as a further complicating factor. It
4
CH3
II
III
is clear then that the validity of the concept of "product development
control" involving the reaction of LiA1H4 or methyl magnesium bromide
(Grignard reagents) as well as other reducing and alkylating reagents
with ketones can be more rigorously tested using this system.
Purpose
In an attempt to broaden the scope of this study, three series of
reagents were studied: (1) LiBH4 , LiAlB4 and LiGaH4 , (2) BH3 , AlH3 and
GaB3 , and (3) (CH3 ) 2Be, (CH3 ) 2Zn, (CH3 ) 2Ng and (CH3 ) 3A1. These reagents
were allowed to react with 7-norbornanone (I), exo-2-methyl-7-norbornanone
(II) and endo-2-methy1-7-norbornanone individually and competitively in
order to evaluate the importance of "product development control" in the
reduction and alkylation of ketones compared to "steric approach control".
5
CHAPTER II
EXPERIMENTAL
General Considerations
Manipulations of air-sensitive compounds were performed under
nitrogen in a glove box equipped with a recirculating system using man-
ganeous oxide columns to remove oxygen and dry ice-acetone traps to re-
move solvent vapors.16 R
eactions were performed under nitrogen or
argon at the bench using Schlenk tube techniques.17
Calibrated syringes
equipped with stainless steel needles were used for transfer of reagents.
Glassware and syringes were flamed or heated in an oven and cooled under
a flow of nitrogen or argon. All standard solutions were prepared by
weighing the solute in a tared volumetric flask and diluting with the
appropriate solvent.
Materials
Fisher reagent grade anhydrous diethyl ether was distilled under
nitrogen from LiA1H4 prior to use. Fisher reagent grade tetrahydrofuran
(THE) was distilled under nitrogen from NaA1H4 prior to use. Hexachloro -
cyclopentadiene was obtained from Aldrich Chemical Conpany and used
without further purification.
Lithium and sodium aluminum hydride solutions were prepared by
refluxing LiA1H4 (Alfa Inorganics) in THE or Et 20 and NaA1H4 (Alfa
Inorganics) in THE for at least 24 hours followed by filtration through
a fritted glass funnel using dried Celite as a filter aid in the dry
12 box. The clear solution was standardized for aluminum content by EDTA..
The same procedure was followed for LiBH4
(Metal Hydrides).
Gallium chloride was purchased from Alfa Inorganics and used
without further purification.
Borane in THE was purchased from Alfa Inorganics. Before using,
the ratio of borane to hydride was checked.
The alkylating agents, Me2Be, Me2Zn and Me 2Mg were obtained from
co-worker Dr. H. S. Prasad. Trimethylaluminum was obtained from the Ethyl
Corporation. Before use, it was distilled under vacuum in the dry box.
Apparatus
All melting pdints are corrected and all boiling points are
uncorrected. The proton NMR spectra were determined at 60 MHz with a
Varian, Model A-60 or Model T-60 spectrometer. The carbon-13 NMR spectra
were obtained at 125 MHz with a JEOL Fourier transform spectrometer, Model
PFT-100. The chemical shift values are expressed in b values (ppm)
relative to a Me4Si internal standard. The mass spectra were obtained
with a Hitachi (Perkin-Elmer), Model RMU-7 or a Varian, Model M-66, mass
spectrometer. GLPC analyses were carried out on an F and M Model 700 or
Model 720 gas chromatograph. The it spectra were determined with a
Perkin-Elmer, Model 621 or Model 257, infrared recording spectrophoto-
meter. Flame photometry was conducted on a Coleman Model 21 photo meter.
Analytical
Gas analyses were carried out by hydrolyzing samples with hydro-
chloric acid on a standard vacuum line equipped with a Toepler pump.17
Magnesium was determined by titrating hydrolyzed samples with standard
EDTA solution at pH 10 using Eriochrome-Black T as an indicator.
Aluminum was determined by adding excess standard EDTA solution to
hydrolyzed samples and then back titrating with standard zinc acetate
solution at pH 4 using dithizone as an indicator. Lithium reagents
were analyzed by the standard Gilman double titration method (titration
of total base then titration of total base after reaction with benzyl
chloride).18
Lithium was also determined by flame photometry. Halide
was determined by titration with AgNO 3 and back titration by KCNS with
ferric alum indicator. The amount of active C-Mg and C-Li was determined
by titrating the active reagent with dry 2-butanol in xylene using 2,2'-
diquinoline as an indicator. Boron analyses were accomplished by the
titration of boric acid-mannitol with standard ra0H. 19 Carbon, hydrogen
analyses were carried out by Atlantic Microlab, Inc., Atlanta, Georgia.
Preparations
Alane
Alane (A1H3 ) was prepared by the reaction of 100% H 2SO4 with
LiA1H4
in TEF at low temperature (dry ice-acetone temperature) and
filtered in the dry box.20
Analysis: Li:Al:H = 0.02:1.0:3.0.
Lithium Gallium Hydride
Lithium gallium hydride (LiGaH4 ) was prepared according to a
modification of a reaction reported in the literature. 21 The apparatus
used was a 1000 ml. three-necked ground glass joint round bottom flask.
An equalizing pressure addition funnel was fixed to the flask. A rubber
septum cap was fixed to the top of the addition funnel. One joint of the
three-necked , flask was plugged and a one-way valve was fastened to the
other joint of the three-necked flask which in turn was connected to the
vacuum manifold and dry'nitrogen so that an inert atmosphere of dry nitro-
gen was maintained.
In a typical synthesis 98.82g (0.56 mole) of gallium (III)
chloride was dissolved in anhydrous diethyl ether. The flask was cooled'
with dry ice-acetone mixture because the dissolving of gallium (III)
chloride in diethyl ether was very exothermic. A 100% excess of lithium
hydride (36.64 g, 4.5 mole), obtained from Alfa Inorganics, and a stirring
bar were placed in the three-necked round bottom flask under dry nitrogen
after which 50 ml of anhydrous diethyl ether was syringed into the flask.
The flask was cooled to -78 °C under constant stirring. Next the gallium
(III) chloride-ether solution was syringed into the addition funnel con-
nected to the three-necked flask. The gallium (III) chloride-ether solu-
tion was added dropwise to the cold, stirred lithium hydride slurry. Slow
addition of gallium chloride was important so that the solution did not
heat up and cause thermal decomposition of the product (lithium gallium
hydride). After the complete addition of gallium chloride-ether solution,
the lithium gallium hydride solution was allowed to warm to 0 °C. The
solution was filtered and the precipitate washed with diethyl ether to
maximize yield. The reaction for the procedure was:
(C 2 H5 2 ) 0 GaC1
3 + 4LiH
L iGaH4 + 3L iC1
The analysis22 showed a hydride to gallium ratio of 3.93 to 1.00
compared to the predicted 4.00 to 1.00 ratio. The per cent yield was
estimated by comparing volumes of the ether solution and the volume of
9
water in a similar flask. Approximately 810 ml. of a 0.61 molar LiGaH 4 .
was produced. The per cent yield was 89%.
The solution was stored in a freezer under-dry nitrogen. After
60 days no noticeable decomposition was observed. After this time the
solution was reanalyzed and the hydride to gallium ratio was 4.00 to
1.00 and the concentration of gallium was 0.63 M. Lithium analysis was
accomplished by flame photometry. The lithium: gallium: hydride ratio
was 1.08 to 1.00 to 4.00. The solution was found to be chloride free
after hydrolysis.
Gallane
Gallane (GaH3 ) was prepared by the same procedure used to prepare
alane. At -78°C, 100% H2SO4
was added slowly to the calculated volune
of LiGaH4
in Et20. After the desired time with stirring, the reaction
mixture was allowed to warm to room temperature and filtered by the
usual manner in the dry box. Analysis: Li:Ga:H = 0.3:1.0:3.0.
5,5-Dimethoxy-1,2,3,4-tetrachlorocyclopentadiene (IV)
Hexachlorocyclopentadiene was converted to (IV) with methanolic
potassium hydroxide according to a modified procedure by McBee
.23
In a
3-1. three-necked round bottom flask fitted with a condenser, an addition
funnel, and a mechanical stirrer were placed 254 g (0.93 mole) of hexa-
chlorocyclopentadiene and 800 ml. of commercial grade methanol. The
stirrer was started, and a solution of 120 g (2.14 mole) of potassium
hydroxide in 600 ml. of methanol added dropwise over a period of
three hours. The reaction mixture was stirred for an additional two
hours and then poured over 3 1. of crushed ice. After the ice had
melted, the mixture was extracted with three 250-ml portions of dichloro-
methane. The combined extracts were dried over anhydrous magnesium
sulfate and concentrated to a yellow syrup on a rotary evaporator. The
residue after distilling through a 12 inch Vigreux column yielded 188 g.
(76%) of 5,5-dimethoxy-1,2,3,4-tetrachlorocyclopentadiene (IV) as a
viscous, yellow-tinted oil, b.p. 79-82 °C (0.6 mm ); NMR (CDC13, TMS)
singlet at 3.30 ppm; mass spectrum, uie (rel. intensity) 237(14 +-37, 2),
236(2), 235(4), 234(5), 233(3), 232(10), 231(2), 230(25), 229(10), 228
(90), 227(15), 226(100), 225(3),224(10), 206(3), 154(20), 58(30); IR
(neat, film) 2970(m), 2940(s), 1610(s), 1450(s), 1190(s), 1140(s), 1100
(m), 1080(m); n25
D 1.5284 [lit.23 b.p. 108-110 (11 mm), n25D 1.5288].
7,7-Dimethoxy-1,2,3,4-tetrachlorobicyclo[2.2.1]hept-2-ene (V)
A 1-1. two-necked round bottom flask was fitted with a fritted-
glass gas inlet tube and a condenser connected to a mercury filled
bubbler. Into the flask was placed 188 g. (0.72 mole) of (IV), and a
slow stream of nitrogen and ethylene was passed through the fritted-
glass inlet tube at a rate to maintain about 1-2 inches of foam in the
flask at 190° C which is heated by an oil bath. The color of the
liquid changed from yellow to redish brown as ethylene was bubbled
through the reaction mixture ar this temperature for 6 hours. The
reaction mixture was cooled and distilled through a 12-inch Vigreux
column to yield 160 g. (75%) of a yellow syrup, b.p. 70-75 (0.15mm);
NMR (CDC13'
TMS) 3H singlet at 3.50 ppm, 3H singlet at 3.55 ppm, 4H
multiplet at 1.63-2.50 ppm; mass spectrum, Jae (rel. intensity) 265(M+-37,
10), 264(2), 263(4), 262(5), 261(4), 260(12), 259(4), 258(28), 257(9),
256(81), 255(15), 254(100), 253(7), 252(14), 234(5), 232(8), 230(9),
228(12), 226(16), 182(20), 180(23), 143(23), 142(18), 133(34), 58(40);
10
11
IR (neat, film) 2970(m), 2940(s), 2830(s), 1610(s), 1460(s), 1290(s),
1270(s), 1220(s), 1190(s), 1125(s), 1095(s), 1060(m), 1015(s), 995(s),
920(s), 875(s), 840(s), 800(s), 735(m); n25D 1.5250 [lit. 24 b.p. 56°C
(0.05 mm), n25D 1.5248].
7,7-Dimethoxybicyclo[2.2.1]heptene (VI)
A 500 ml. three-necked round bottom flask was equipped with a
mechanical stirrer, a condenser fitted with a nitrogen inlet tube to
maintain a slight positive pressure, and a pressure-equalizing dropping
funnel. The flask was placed in a heating mantle, and into it were placed
250 ml. of THE which had been distilled from NaA1H 4, 23 g. (1 mole) of
sodium metal chopped into 5-mm. cubes and 30 ml. (25 g., 0.33 mile) of
t-butyl alcohol. This mixture was stirred vigorously and brought to
gentle reflux. As soon as refluxing occurred, 17.7 g (0.06 mole) of (V)
was added dropwise over a 2 hour period. The mixture was heated under
reflux for 36 hours, cooled to room temperature and the excess sodium
was destroyed by slow addition of 100 ml. methanol. The reaction mix-
ture was poured over 2 1. of crushed ice and extracted with 200 ml. of
diethyl ether. The aqueous phase was separated and the organic phase was
washed with 3-250 ml. portions of saturated aqueous sodium chloride.
The ethere a l solution was dried over anhydrous magnesium sulfate and
concentrated to a dark oil by removal of the ether by a rotary evapo-
rator. The residue was distilled through a 12-inch Vigreux column to
yield 6.1 g. (65%) of colorless liquid, b.p. 60-63 °C (17 nm); IR (neat,
MO 3060 (s), 2950 (s), 2820 (m), 1620 (s), 1190 (m), 1140 (m), 1105
(m), 1080 (n); NMR (CDC1 3 , THS) 2H multiplet at 0.75- 1.01 ppm, 2H
multiplet at 1.63-2.01 ppm, 2H multiplet at 2.59-2.75 ppm, 3H singlet
12
at 3.05 ppm, 3H singlet at 3.13 ppm, 2H multiplet 5.90-6.07 ppm; mass
spectrum, m/e. (rel. intensity) 154(M+, 58), 139(18), 123(58), 107(22),
91(40), 79(100), 59(32); n25 p 1.4645 [lit. 25 b.p. 74-78 °C (30 mm);
n28.6D 1.4584].
Anal. Calcd for C9 H14 02
: C, 70.10; H, 9.15. Found: C, 70.19;
H, 9.13.
This reaction proceeded smoothly when the listed quantities were
used; however, when this reaction was scaled up 8 fold, a violent ex-
plosion ensued. Therefore caution should be exercised.
7,7-Dimethoxy[2.2.1 ]heptane (VII)
In a hydrogenation flask equipped with a magnetic stirrer, 6.1
of (VI) and 0.03 g. of 5% palladium on carbon were stirred under
hydrogen at room temperature and atmosphere pressure. After the hydrogen
absorption had ceased, the catalyst was removed by filtration. The
filtrate was vacuum_ distilled to give 5.5 g. (87%) of 7,7-dimethoxybi-
cyclo[2.2.1]heptane, b.p. 78-80°C (30mm); NMR (CDC1 3, TMS) 10H multiplet
at 0.83-2.04 ppm, 6H singlet at 3.18 ppm; IR (neat, film) 2980(s),
2960(s), 1450(s), 1195(m), 1130 (m), 1100(m), 1050(m); mass spectrum, mje
(rel. intensity) 156(1+, 100), 125(90), 111(33), 94(40), 83(88), 55(87);
n25D 1.4530 [lit. 25 b.p. 80°C (30 mm); n25D 1.4533].
Anal. Calcd for C 9H1602 : C, 69.19; H, 10.32. Found: C, 69.46;
H, 10.28.
Bicyclo[2.2.1]heptan-7-one (I)
In a 25-m1. round bottom flask with distillation head, 2.06 g.
(0.015 mole) of (VII) and 15 ml. (0.25 mole) of glacial acetic acid were
heated to 115°C for 10 hours. After cooling, the solution was trans-
13
ferred to a separatory funnel with 30 ml. of petroleum ether (b. p. 35- •
45°C). A solution of 12 g. (0.30 mole) of sodium hydroxide in 40 ml. of
water was carefully added dropwise with cooling. An additional 40 ml. of
water was added to dissolve the sodium acetate that had formed. The water
layer was separated and washed twice with 25 ml. portions of petroleum
ether. The combined petroleum ether extracts were dried over anhydrous
magnesium sulfate, and the drying agent was removed by filtration. The
filtrate was concentrated to 15 ml. The temperature was lowered to -20 °
to initate crystallization. After crystallization was complete, the
solvent was removed and the precipitate dried by passing nitrogen over
the crystals to give 0.90 g. (62%) of bicyclo[2.2.1]heptan-7-one (I)
[7-norbornanone], m.p. 77-79 °C; IR (Nujol) 1845(w), 1760(s), 1740(w),
2920(s),2840(s), 1460(s), 1380(n), 1300(m), 1200(m), 1180(w), 1160(m);
NKR (CDCL3 , TMS) multiplet at 1.50-2.20 ppm; rass spectrum, m/e (rel.
intensity) 110(M+, 80), 81(75), 67(95), 55(100);[lit. 25 m.p. 79.5-80.5°C;
IR. 1754(s), 1783(w)].
Anal. Calcd for C H100: C, 76.32; H, 9.15. Found: C, 76.40;
H, 9.32.
Exo-7, 7-dinethoxynorbornan-2-ol (ATM)
A 500-mi. round bottom flask equipped with a magnetic stirrer
was charged with 12.76 g (0.04 mole) of mercuric acetate followed by 40
ml. of water. The mercuric acetate dissolved to give a clear solution.
Tetrahydrofuran (40 ml) was added to the solution forming a bright
yellow suspension. To the mixture, which was cooled in an ice bath,
was added 0.038 mole of (VI) with stirring. The mixture was allowed to
warm to room temperature with stirring until the reaction mixture became
14
colorless and clear (30 minutes). Stirring was continued for an addi-
tional 15 minutes and 40 ml. of 3 N sodium hydroxide was added followed
by 40 ml. of 0.5 M sodium borohydride in 3 M sodium hydroxide. The
reduction was almost instantaneous. The mercury was then allowed to
settle and the water layer saturated with sodium chloride. The tetra-
hydrofuran layer was separated, dried with anhydrous sodium - sulfate and
filtered after which the solvent was removed by rotary evaporation.
The crude product was distilled to give 0.030 mole, (5.2 g., 80%), of
(VIII)b.p.88-91°C0-7=0;2633 1.4372: IR(neat, film) 3600-3200(broad-s)
2950(s), 1170(0,1145(m), 1115(m),1060(m); NMR (CDC1 3 , TMS) 6H multiplet
at 0.88-2.17 ppm, 1H broad singlet at 2.92 ppm, 1H multiplet at 3.10 ppm,
3H singlet at 3.18 ppm, 3H singlet at 3.24 ppm, 2H multiplet at 3.34-
3.95 ppm; mass spectrum, m/e (rel. intensity) 172(14+,0,154(44), 141(24),
117(26), 115(22), 111(29), 109(22),108(29), 101(100), 97(64), 96(41),
91(64), 81(38), 80(38), 79(30), 67(20), 55(40).
Anal. Calcd for C9H1603 : C, 62.76; H, 9.37. Found: C, 62.55;
H, 9.29.
7,7-Dimethoxynorbornan-2-one (IX)
A 1-1. three-necked round bottom flask was equipped with a magnetic
stirrer, a thermometer immersed in the reaction mixture and a nitrogen
inlet tube. In the flask was placed 100 ml of anhydrous pyridine, and
the flask was cooled in an ice-water bath to 15-20 °C. Chromium trioxide
(10.0 g. ) was added in small amounts to the stirred solvent at a rate so
as to keep the temperature below 30 °C. After about one-third of the
Cr03
was added, the yellow complex began to precipitate. At the end of
the addition (about 1 hour), a slurry of the yellow-orange complex in
15
pyridine remained. The temperature of the stirred solution was readjusted
to 15°C., and the stirring at this temperature was continued until the
precipitate reverted to a deep red macrocrystalline form. Petroleum
ether (200 ml.) was then added to the reaction mixture, the precipitate
was allowed to settle, and the solvent mixture was decanted. The residue
was washed three times with 200 ml of petroleum ether with the solvent
being removed each time by decantation. The precipitate was dried by
passing nitrogen over it. Dry methylene chloride was added to make a 5%
solution. A solution of 16.5 g. (0.096 mole) of (VIII) in 20 ml. of dry
methylene chloride was added with stirring to the chromium trioxide-
pyridine complex at room temperature. The oxidation was complete in
about 15 minutes with a black precipitate forming. The solvent was
removed by a rotary evaporator leaving the crude product which was
distilled to yield 13.3 g. (0.078 mole, 82%) of (IX), b.p. 71-72 °C(1-2=);
m.p. 32-33 °C; IR(CDC13 , cavity cell) 2950(s), 1760(s), 1200(m), 1150 (m),
1110(m), 1080(m); NMR (CDC13 , TMS) 6H multiplet at 1.27-2.57 ppm, 3H
singlet at 3.23 ppm, 3H singlet at 3.26 ppm, 2H multiplet at 3.07-3.30 ppm;
mass spectrum, m/e (rel. intensity) 170(e, 14), 155(2), 139(10), 123(4),
115(8), 101(100), 97(34), 67(8), 59(8), 55(32), 41(18), 39(12).
Anal. Calcd for C9 H14 03.• C, 63.51; H, 8.29. Found: C, 63.62;
H,
2-Methylene-7,.77dimethoxynorbornane (X)
A solution of sodium methylsulfinyl carbanion was prepared
according to the procedure of Corey.26
Into a three-necked, 500 ml.
round bottom flask, was placed 3.84 g. (0.08 mole) sodium hydride (50%
mineral oil dispersion). The sodium hydride was washed three times with
16
petroleum ether by swirling allowing the hydride to settle, and decanting
the liquid portion in order to remove the mineral oil. The flask was
immediately fitted with a magnetic stirrer, a reflux condenser, and a
pressure-equalizing dropping funnel. A three way stopcock, connected to
the top of the reflux condenser, was connected to a water aspirator and
a source of dry nitrogen. The system was evacuated until the last traces
of petroleum ether were removed from the sodium hydride and was then
flushed with nitrogen by evacuating and filling the nitrogen several
times. The aspirator hose was removed and this arm of the stopcock was
connected to a bubbler to which the system is opened. Dimethyl sulfoxide
which was distilled from calcium hydride (b.p. 64 °/4 mm) was introduced
through the dropping funnel and the mixture was heated with stirring to
70-75°C until the evolution of hydrogen ceases which usually was about
45 minutes. The solution was cooled in a cold water bath and stirred
during the addition of 17.8 g. (0.077 mole) of (methyl)-triphenylphos-
phonium bromide in 50 ml. of warm dimethyl sulfoxide whereupon the deep
red color of the ylide was produced. After stirring for 15 minutes the
ketone (IX) in 10 ml. of dimethyl sulfoxide was added with stirring in
a cold water bath. The reaction mixture was heated to 60 ° for four hours.
The reaction mixture was then cooled and poured into 500 ml. of cold
water. The mixture was extracted three times with pentane, washed once
with water, dried over anhydrous sodium sulfate, filtered and the solvent
removed by rotary evaporation. The crude product was distilled to pro-
vide 10.0 g (0.06 mole, 77%) of (X) , b.p. 76-80°C (15 mm); n261) 1.4575;
IR(neat, film) 3060(w), 2940(s), 2820(m), 1670(m), 1325(s), 1280(m),
1200(m), 1170(m), 1130(m), 1100(m), 1080(s), 880(s); NHR(CDC1 3 , TATS) 8H
17
multiplet at 0.83-2.67 . ppm, 3H singlet at 3.16 ppm, 3H singlet at 3.18 ppm,
2H multiplet at 4.51-4.83 ppm; mass spectrum, m/e (rel. intensity) 168(M+ ,
60), 153(15), 137(60), 123(13), 121(25), 105(40), 95(12), 93(100), 91(20),
79(30), 77(10), 75(10), 59(32).
Anal. Calcd for C 10H160 2 : C, 71.39; H, 9.59. Found: C, 71.25;
H, 9.69.
2-Methy1-7,7-dimethoxynorbornane (XI)
In a hydrogenation flask, CO was added to 10 ml. of 95% ethanol
and 0.3 g. of 5% palladium on carbon. This mixture was stirred under
hydrogen at room temperature until the amount of hydrogen (1344 ml.) had
been taken up. The catalyst was removed by filtration, and the crude
product was purified and the exo- and endo-isomers were separated by
preparative gas chromatography using a 6-foot, 1/2 inch inner diameter,
20% carbowax 20 M on chromosorb W-NAW at 125 °C with a flow rate of 6.5
1 cm
3 /minutes. The crude products' ir and H nmr matched a similar mixture
of the glc collected isomers' ir and 1H nmr spectra, thus insuring that
no decomposition on the column was taking place. The exo-ketal (XIa)
was in a ratio of 7:1 with the endo-ketal ((Ib). The first compound
eluted was identified as endo-2-methyl-7,7-dimethoxynorbornane (KI 13 );
n25D 1.4501; IR(neat, film) 2940(s), 2860(m), 2820(s), 1450(s), 1380(m),
1325(s), 1270(m), 1200(s), 1140(s), 1100(s), 1080(s), 1060(s), 1030(m),
1000(s); NMR (CDC13, TMS) 3H doublet at 0.95 ppm, J = 6 Hz, 9H multiplet
at 1.17-2.34 ppm, 6H singlet at 3.17 ppm; mass spectrum, m/e (rel. inten-
sity) 170(M+, 100), 155(97), 139(59), 129(42), 115(89), 101(97), 9 7 (66 ),
55(63).
Anal. Calcd for C10 H18 02.• C, 70.55; H, 10.66. Found: C, 70.42;
H, 10.62.
18
The second compound eluted was identified as exo-2-methyl- 7,7-di-
methoxynorbornane (XIa); n25 1.4320; IR(neat, film) 2950(s), 2860(m),
2820(s), 1450(s), 1330(s), 1270(m), 1200(s), 1140(s), 1100(s), 1075(s),
1055(s), 1030(m), 1000(s); NMR (CDC13 , TMS) 3H doublet at 1.05 ppm, J =
6 Hz, 9H multiplet at 0.80-2.05 ppm, 3H singlet at 3.16 ppm, 3H singlet
at 3.17 ppm; mass spectrum, m/e (rel. intensity) 170(M+, 100), 155(75),
139(60), 138(20), 129(50), 115(80), 101(90), 97(70), 55(70).
Anal. Calcd for C10H1802 : C, 70.55; H, 10.66. Found: C, 70.51;
H, 10.60.
Exo-2-methyl-7-norbornanone (II)
Into a 50 ml Erlenmeyer flask equipped with a magnetic stirrer,
the exo-ketal (XIa) (2.50 g, 14.7 mmole) was added to 25 ml of 5% H2SO
4
and stirred for 12 hours at room temperature. Afterwards, this mixture
and 25 ml of diethyl ether were placed into a separatory funnel. The
ether layer was separated and the aqueous layer was washed twice with di-
ethyl ether. The combined ethereal fractions were then washed with water,
saturated sodium bicarbonate, water, saturated sodium chloride and dried
over anhydrous magnesiun sulfate. The drying agent was removed by filtra-
tion. The ether was reduced to a volumn of 20 ml by a rotary evaporator.
The crude ketone was purified by preparative gas chromatography on a 10-
foot, 1/4 inch inner diameter column, 20% Carbowax 20i on Chromosorp W-NAW
at 125°C with a flow of 30 ml of He/min. No starting ketal was observed
in the glc trace. With these conditions, the ketal would have had a re-
tention time of 10.5 minutes. The product had a retention time of 15.0
minutes. The compound was identified as eKo-2-methyl-7-norbornanone (II);
n25 1.4490; IR(neat, film) 2940(s), 2850(w), 1840(w), 1765(s), 1730( ),
19
1450(m), 1140(s); NMP. (CDC13, TMS) 3E doublet at 0.96 ppm, J = 6 Hz, 9H
multiplet at 1.33-2.17 ppm [lit. 27 NMR (CDC13 , TMS) 3H doublet at 0.96
ppm]; mass spectrum, m/e (rel. intensity) 124(M+, 80), 109(25),, 95(60),
93(50), 81(90), 68(40), 67(50), 55(100).
Anal. Calcd for C 8H1 0: C, 77.37; H, 9.74. Found: C, 77.30;
H, 9.72.
Endo-2-methyl-7-norbornanone (III)
The same procedure used to convert the exo-ketal to the exo-ketone
(II) was used to convert the endo-ketal (XIb) to the endo-ketone (III).
Again, under the same conditions described above for the purification and
isolation of the exo-ketone, no starting ketal (XIb) was observed in the
glc trace. Under these conditions the ketal would appear after 11.5
minutes and the endo-ketone appeared after 16.5 minutes. This ketone was
identified as endo-2-methyl-7-norbornanone (III); n25D
1.4661; IR(neat,
film) 2940(s), 2850(m), 1860(w), 1740(vs), 1470(m), 1450(m), 1145(s); NMR
(CDC13, TMS) 3H doublet at 1.10 ppm, J = 6 Hz, 9H multiplet at 1.62-2.42
ppm [lit. 27 NMR (CDC13 , TMS) 3H doublet at 1.10 ppm]; mass spectrum, mle
(rel. intensity) 124(1+, 83), 109(12), 95(79), 93(34), 81(98), 68(41),
67(52), 55(100).
Anal. Calcd for C812 H 0: C, 77.37; H, 9.74. Found: C, 77.32;
H, 9.70.
5-Methyl-7,7-dimethoxy-1,2,3,4-tetrachloronorborn-2-ene (XII) (1:9 Mixture
of exo- to endo-isomers)
A. mixture of propylene and nitrogen was bubbled in to 264 g.(1
mole) of (IV) under the same conditions used to prepare (V). Compound.
(IV) was preheated to 190°C in a 500 ml. one-necked round bottom flask
20
fitted with a fritted glass gas inlet tube and a magnetic stirrer. Once
the starting material reached a temperature of 190 °C, the gas was intro-
duced and these conditions were maintained for 6 hours with stirring.
After the desired time, the reaction mixture was cooled and distilled to •
provide 230.0 g. (0.75 mole, 75%), of a pale-yellow syrup identified as
(XII); b.p. 88-91° C (0.7 mm) n25D
1.5204; IR(neat, film) 2960(m), 2940(s),
2835(s), 1620(w), 1610(s), 1450(s), 1380(m), 1275(s), 1200(s), 1130(s),
1095(s), 1030(m), 1000(m), 965(m), 910(m), 785(s), 760(s); mass spectrum,
m/e (rel. intensity) 268(1%1+-38,6) 9 267(2), 266(14), 265(4), 264(36),
263(6), 262(38), 261(2), 257(15), 255(30), 253(15), 249(33), 247(24),
239(9), 237(15), 235(36), 233(58), 231(61), 229(100), 227(79), 225(21),
223(33), 221(48), 220(52), 219(39), 218(88), 216(82), 214(27), 212(33),
190(64), 120(61), 118(79), 59(79); NMR (CDC1 3 ,TMS) 3H doublet at 0.92 ppm
J= 6 Hz, 3H multiplet at 1.21-2.75 ppm, 3H singlet at 3.51 ppm, 3H singlet
at 3.56 ppm.
Anal. Calcd for C H 10 1 : C, 39.24; H, 3.95. Found: C, 39.11;
H, 3.89.
A 10% SE-30 on Chromosorb H column at 180°C with a flow rate of 55
ml of He/min providing a glc trace which showed a 1 to 9 ratio of the exo-
and endo-isomers. The isomers were not separated at this stage of the
synthesis.
5-Methyl-7,7-Dinethoxynorborn-2-ene (UII) (A 1:9 Ratio of Exo- to Endo-
Isomers)
To a vigorously stirred solution of 90 g. (1.22 moles) of t-butyl
alcohol, 525 ml. of freshly distilled tetrahydrofuran, and 59 g.(2.57
g-atoms) of finely chopped sodium metal under a nitrogen atmosphere in a
21
1-1 three-necked round bottom flask equipped with a magnetic stirrer, a
pressure-equalizing dropping funnel fitted with a rubber serum cap and a
condenser fitted with a three-way stopcock lonnected to a mineral oil
filled bubbler was added 30.6 g. (0.1 mole) of (XII). The mixture was
heated gently to maintain a steady reflux for 10 hours. After cooling,
the excess sodium was destroyed by slow addition of methanol ( about
500 ml.) to the reaction mixture. The reaction mixture was poured over
2 1. of ice and the reaction flask was washed with approximately 600 ml.
of water. The solution was extracted with three 250 ml. portions of
water and once with saturated sodium chloride solution. The ethereal
solution was dried over anhydrous sodium sulfate and filtered. The glc
confirmed the crude product contained both the exo- and endo-isomers
in a 10 to 90 ratio. The crude product was distilled to give 9.8 g. of
the mixture of isomers ( 59% yield), b.p. 73-78 °C (13 mm); 1125D
1.4601;
IR(neat, film) 3060(m), 2950(s), 2840(m), 2820(m), 1600(m), 1450(s),
1300(s), 730(s); SIR (CDC1 3, TMS) 3H doublet at 0.76 ppm J = 6.5 Hz,
2H multiplet at 0.30-1.38 ppm, 1H multiplet at 1.78-2.32 ppm, 2H mul-
tiplet at 2.47-2.73 ppm, 3H singlet at 3.02 ppm, 3H singlet at 3.12
ppm, 2H multiplet at 5.83-6.20 ppm; mass spectrum, m/e (rel. intensity)
168(M+, 60), 153(46), 137(46), 123(30), 93(100).
Anal,. Calcd for C10H16
02: C
, 71.39; F, 9.59. Found: C, 71.30;
H, 9.50.
Hydrogenation of (XIII)
In a hydrogenation flask, 9.8 g. of (KIII), was added to 10 ml.
of 95% ethanol and 0.10 g. of 5% palladium on carbon. This mixture was
stirred under hydrogen at room temperature. After the hydrogen absorp-
tion had ceased, the catalyst was removed by filtration. The crude
product was purified and the exo- and endo-isomers were separated by — —
preparative glc using a 15 ft. 20% Carbowax 20 M on Chromosorp W-NAW
column at 130°C with a flow of 30 ml/min yielding a 9:1 ratio of endo-
and exo-ketals. The nmr, it and mass spectrum as well as glc retention
times of these compounds were identical to the exo- and endo-ketals,
(XIa) and (XIb), prepared above.
7-Norbornanol (XIV)
Into a 250-m1. three-necked round bottom flask equipped with a
magnetic stirring bar and fitted with a three-way stopcock with one arm
connected to a mineral oil filled bubbler and a pressure-equalizing
dropping funnel fitted with a rubber serumcap,a solution of 1.03 g.
(9.30 mmole) of ketone (I) in 20 ml. of distilled Et 20 was placed.
After cooling the flask and contents to 0° C, a solution of .741 g.
(19 mmole) of LiA1H4 in 40 ml of Et20 was added dropwise with stirring.
After addition was complete the solution was stirred for 2 hours where-
upon 3 ml. of water was added dropwise to the stirring solution of 0°C.
The resulting slurry was stirred for 0.5 hour, filtered, and the
precipitate washed with Et20. The filtrate was dried over anhydrous
magnesium sulfate and the drying agent was removed by filtration. The
ether was removed by flash evaporation to provide 0.897 g. (86%) of
22
23
(XIV), m.p.. 151-153°C. The white crystals were recrystallized from
hexane to give pure (XIV), m.p. 152-153 °C (lit. 28 m.p. 152-153 °C).
Birch Reduction of (II)
To a 100 ml. three-necked round bottom flask equipped with a
magnetic stirrer, dry ice-acetone condenser and a stopper was added 50
ml. of condensed anhydrous ammonia, 0.122 g. (0.001 mole) of (II) and
2.5 ml. of absolute ethanol. To this stirred reaction mixture, 0.7 g.
of finely chopped sodium metal was added. A deep blue color was produced.
Stirring was continued for 15 minutes and the ammonia allowed to evaporate
after the addition of 2.5 ml. of water. The mixture was extracted twice
with hexane. The combined hexane extracts were washed with water and
saturated sodium chloride solution and then dried with anhydrous sodium
sulfate. The hexane was removed by rotary evaporation. The residue
showed two peaks on the gas chromatograph using a 15 ft., 20% carbowax
20 M on chromosorp W-NAW column in a 20:80 ratio.
The first compound, minor product, eluted' was identified as
exo-2-methyl-syn-7-norbornanol, (oa); n D 1.4507; IR(neat, film) 3600-
3000(broad-s), 2940(s), 2860(s), 1450(s), 1380(s), 1360(s), 1340(m),
1310(s), 1170(s), 1160(s), 1110(s), 1090(s), 1050(m), 1120(m), 850(m);
mass spectrum, m/e (rel. intensity) 126(M+, 7), 111(35), 108(63),
97(43), 95(90), 93(55), 91(15), 84(20), 83(20), 82(30), 79(30), 71(25),
70(100), 69(25), 67(60), 57(62), 55(55), 53(15), 43(20), 41(40),
40(30), 39(30); H NMR (CDC13 , THS) 3H doublet at 0.91 ppm J= 6 Hz,
10H multiplet at 1.05-2.15 ppm, 1H singlet at 3.97 ppm; 13C NMR(CDC13'
multiplicity in off-resonance decoupling) C(7), 80.38 ppm (d); C( 1 ),
46.29 ppm (d); C(2), 41.37 ppm (d); C(4), 36.95 ppm (d); C(3), 36.34 ppm
(t); C(6), 27.73 ppm. (9; C(5), 25.60 ppm (t); C(8), 22.33 ppm (q).
Anal. Calcd for C8 H14 '0: C, 76.14; H, 11.L8. Found: C, 76.01;
H, 11.23.
The second compound, major product, was identified as exo-2-
methyl-anti-7-norbornanol, (XVb); n 261) 1.4625; IR(neat, film) 3600-3000
(broad- ), 2940(s), 2860(s), 1450(s), 1380(s), 1360(s), 1340(m), 1320
(s), 1180(s), 1150(s), 1110(s), 1080(s), 1050(m), 1020(m), 840(m);
mass spectrum, mle (rel. intensity) 126(M+, 10), 111(18), 108(33), 97
(16), 95(81), 93(64), 91(14), 84(24), 83(25), 82(34), 79(26), 71(34),
70(100), 69(28), 67(55), 57(50), 55(53), 53(19), 43(26), 41(40), 39(33);
IH NMR (CDC13, TVS) 3H doublet at 1.12 ppm J= 6 Hz, 10H multiplet at
1.05-2.15 ppm, 1H singlet at 4.10 ppm; 13C NMR (CDC13 , multiplicity in
off-resonance decoupling) C(7), 77.90 ppm (d); C(1), 46.66 ppm (d); C(2),
41.07 ppm (d); C(4), 36.77 ppm (1); C(3), 34.89 ppm (t); C(6), 27.06 ppm
(t); C(5), 29.60 ppm (t); C(8), 22.02 ppm (q).
Anal. Calcd for C 0: C, 76.14; H, 11.18. Found: C, 76.05;
H, 11.25.
Endo-2-methyl-7-Norbornanol (XVI)
The same procedure used to prepare 7-norbornanol (XIV) was followed
to prepare the syn- and anti-endo-2-methyl-7-norbornanols (XVa) and (XVb),
respectively. A solution of ketone (III) in Et 20 was added dropwise
to the calculated amount of LAH in Et 20. After the normal work up, the
crude alcohols were obtained in an Et 0 solution. This solution was 2
subjected to the aforementioned glc conditions and only one ..peak was
observed which was expected since the steric environment for both
isomers is essentially the same. These isomers were collected together
24
25
via preparative glc. The collected mixture had the following properties:
m.p. 51-59°C; IR (CDC13 , Cavity Cell) 3600(sharp-s), 3550-3050(broad-s),
1480(s), 1450(s), 1380(s), 1350(s), 1300(s), 1260(m), 1230(w), 1175(w),
1075(s), 1050(w); mass spectrum, pile (rel. intensity) 126(M+, 7), 111(17),
109(43), 97(29), 95(100), 93(95), 85(14), 84(24), 83(29), 82(29), 81(14),
80(24), 79(29), 77(19), 71(43), 70(100), 69(33), 68(24), 67(38), 57(38),
43(38), 41(57), 40(71), 39(38); 1H NMR (CDC13 , TMS) 3H doublet at 1.02
ppm J = 7 Hz, 9H multiplet at 0.53-2.50 ppm, 1H singlet at 2.48 ppm, 1H
broad singlet at 4.08 ppm. The 13C NMR was conducted in order to confirm
the existence of both isomers. Based on literature values2 9 the following
assignments were made for endo-2-methyl-syn-7-norbornanol (XVIa) and endo-
2-methyl-anti-7-norbornanol (XVIb): (XVIa) 13C NMR (CDC13, multiplicity in
off-resonance decoupling) C(7), 81.05 ppm (d); C(I), 45.38 ppm (d); C(2),
41.32 ppm (d); C(4), 35.67 ppm (d); C(3), 29.91 ppm (0; C(6), 27.24 ppm
(t); C(5), 19.54 ppm (0; C(8), 16.44 ppm (q). (XVIb) 13C NMR (CDC1 3 ,
multiplicity in off-resonance decoupling) C(7), 80.02 ppm (d); C(1),
45.75 ppm (d); C(2), 41.62 ppm (d); C(4), 35.98 ppm (d); C(3), 30.34 ppm
(0; C(6), 27.29 ppm (0; C(5), 19.17 ppm (0; C(8), 17.29 ppm (q). The
ratio of (XVIa) to (XVIb) was determined to be approximately 50:50 by
comparing the intensities of the 13C NMR lines associated with the C(7)s
and C(8)s of each compound (see eqs. 3 and 4).
Anal. Calcd for C8 H14 0: C, 76.14; H, 11.18. Found: C, 76.20;
H, 11.23.
Reductions of Ketones
A 5 ml Erlemdeyer flask equipped with a magnetic stirring bar, was
flash flamed, cooled and fitted with a rubber serum cap under nitrogen.
26
To the flask was added the calculated amount of each ketone in THE or
Et20 along with the internal standard which was hexadecane. The calcu-
lated amount of reducing agent was then added to the stirred mixture at
the desired temperature. After the desired time, usually two hours, the
reaction was quenched with water or saturated ammonium chloride solution.
The organic layer was separated and dried over anhydrous sodium sulfate and
then subjected to glc conditions for identification of products. A 15 ft.
20% Carbowax 20 M on Chromosorb W-NAW column at 135 °C with a fIow rate of
30 ml of He/min was used to effect separation of all products. Inverse
addition of reactants provided the same results. The following was the
order of elution of the products on this column at these conditions:
ketone (I), 12.0 minutes; ketone (II), 20.0 minutes; ketone (III), 25.2
minutes; alcohol (XIV), 26.3 minutes; alcohol (XVa), 33.0 minutes;
alcohols (XVIa) and (XVIb), 38.3 minutes; alcohol (XVb), 41.4 minutes;
and hexadecane, 48.0 minutes. In all cases, only alcohol (XVa) was pro-
duced when ketone (II) was reduced. As noted before, the isomers of
alcohol (XVI) could not be separated by any technique but could be identi-
fied by 13
C NMR.
Meerwein-Ponndorf-Verley Reduction of Ketone (II)
Into a 50 ml one-necked round bottom flask fitted with a partial
reflux head and a napetic stirrer was placed 2 ml of 0.032 M solution
of ketone (II) in diethyl ether along with 1.0 g aluminum isopropoxide
and 5 ml of isopropyl alcohol. This mixture was heated to 50 °C and
stirred for 2 days. After cooling to room temperature, the mixture was
poured into 100 ml of saturated aqueous ammonium chloride and the
solution extracted with two 10 ml portions of ether. The ethereal
27
extracts were combined and washed with water and saturated aqueous
sodium chloride and dried over anhydrous sodium sulfate. The ether
was partially removed by use of a water aspirator. This solution was
then analyzed as before by glc. The only product was the anti-alcohol.
Meerwein-Ponndorf Equilibration of (XVa) and (XVb)
Into a 50 ml one-necked round bottom flask equipped with a magnetic
stirrer was added 100 mg of syn•2-exo-methyl-7-norbornanol (XVa), 1 g of
aluminum isopropoxide, 5 ml of isopropanol and 5 m1 of acetone. Mild
heat was applied with stirring for 3 days. The glc analysis of the
hydrolyzed mixture was shown to contain the anti-alcohol almost ex-
clusively except for a trace of the syn-alcohol.
7-Methyl-7-Norbornanol (XVII)
Into a 250 ml three-necked round bottom flask equipped the same
as for the preparation of 7-norbornanol (XIV), was placed a solution
of 1.05 g (9.31 mnole) of ketone (I) in 20 ml of freshly distilled Et 20.
To this mixture, a solution of 40 ml of 0.51 M NeVgBr ( 20.4 mnole) in
Et20 is added dropwise with stirring. After the addition was complete,
the solution was stirred for 2 hours and then quenched and worked up in
the usual manner. The ether was removed by rotary evaporation to provide
1.02 g (87%) of white crystals which were sublimed at 70-80°C (15mm) to
give pure (XVII), m.p. 95-96 °C; IR(CC14 , cavity cell) 3625 cm 1 [lit.30
m.p. 97-98°C, IR (CCli , cavity cell) 3618 cm 1 ].
Syn-2-Exo-Methyl-7-Methy1-7-Norbornanol (XVIIIa)
This alcohol was prepared by the sane procedure used to prepare
alcohol (XVII). Ketone (II) was allowed to react with methylmagnesium
bromide which was prepared by the reaction of magnesium metal with
28
methylbromide in Et20 in a Grignard to ketone ratio of 2;1. After the
addition, the resulting solution was stirred for two hours and then
quenched and worked up in the usual manner. The product was then
analyzed on a IO ft. 20% FFAP on Diatoport S column at 150 °C and flow
rate of 30 ml of He/min. Only one compound was observed at these con-
ditions. The product was identified as'syn-2-exo-nethyl-7-methyl-7-
norbornanol (XVIIIa); m.p. 51-52 °C; IR (CDC13, Cavity Cell) 3600(sharp-m),
2950(s), 2870(m), 1480(m), I450(m), 1380(s), I315(w), 1300(w), I265(w),
I225(m), 1200(w), 1185(w), 1170(m), II10(s), 960(m), 950(s); mass spectrum,
m/e (rel. intensity) 140(M+, 7), 125(52), 122(15), I11(9), 107(7), 97(33),
93(14), 85(24), 84(17), 83(10), 82(24), 81(19), 71(57), 69(14), 67(21),
55(53), 43(100), 41(24), 39(17); 1H NMR (CDC13 , TMS) 3H doublet at 1.21
ppm J = 6 Hz, 3H singlet at 1.35 ppm, 10H multiplet at 0.95-1.83 ppm;
13C NMR (CDC13' multiplicity in off-resonance decoupling) C(7), 84.15 ppm
(d); C(I), 49.45 ppm (d); C(2) 44.96 ppm (d); C(4), 38.59 ppm (d); C(3),
38.58 ppm (t); C(6), 29.18 ppm (t); C(5), 26.63 ppm (t); C(9), 22.51 ppm
(q); C(8), 22.14 ppm (q).
Anal. Calcd for C9H 0: C, 77.08; H, 11.50. Found: C, 76.91;
H, 11.59.
Anti-2-Exo-Methyl-7-Methy1-7-Norbornanol (KVIIIb)
Into a 50 ml one-necked round bottom flask fitted with a magnetic
stirring bar, 25 ml of 20% H2 SO4 and 2 mmole (0.28g) of (XVIIIa). This
reaction mixture was stirred for 2 days at 80 °C. Afterwards, the solution
was cooled and 20 ml of Et20 was added. The ethereal layer was separated,
washed with water, saturated sodium chloride and dried over anhydrous
sodium sulfate. Subjecting the product to the same glc conditions used
29
for the syn-alcohol, two compounds were observed. The first compound
eluted matched the retention time for the syn-alcohol and was in a 2:1
ratio with the other compound which was identified as the anti-alcohol
(XVIIIb); m.p. 49-50°C; IR (CDC11 , Cavity Cell) 3590(w), 3330(broad-s),
2950(s), 2860(s), 1485(w), 1450(0, 1380(s), 1350(m), 1300(m), 1245(m),
1220(m), 1190(m), 1130(s), 1100(s), 950(s); mass spectrum, m/e (rel.
intensity) 140(e, 10), 125(35), 122(1), 111(15), 107(10), 97(50), 93(20),
85(15), 84(10), 82(15), 81(23), 71(60), 69(15), 67(20), 55(70), 43(100),
41(20), 39(25); 1H NMR (CDC13, TMS) 3H doublet at 1.07 ppm J = 7 Hz, 3H
singlet at 1.46 ppm, 10H multiplet at 0,96-2.17 ppm; 13C NMR (CDC1 3'
multiplicity in off-resonance decoupling) C(7), 83.06 ppm (d); C(1),
49.69 ppm (d); C(2), 44.90 ppm (d); C(4), 37.68 ppm (d); C(3), 36.16 ppm
(t); C(6), 30.52 ppm (t); C(5), 27.30 ppm (t); C(9), 22.57 ppm (q); C(8),
21.42 ppm (q).
Anal. Calcd for C9H160: C, 77.08; H, 11.50. Found: C, 77.20;
H, 11.35.
' The first compound eluted which matched the retention time of the
syn-alcohol was also isolated and it was indeed confirmed to be the syn-
alcohol by its ir, nmr and mass spectra.
Endo-2-Methy1-7-Methy1-7-Norbornanol (XIX)
The same procedure used to prepare 7-methyl-7-norbornanol (XVII)
was followed to prepare the syn- and anti-endo-alcohols (KIXa) and
(XIXb). A solution of ketone (III) in freshly distilled diethyl ether
was added dropwise to a 100% excess of lieMgBr in diethyl ether. After the
normRT reaction time, the usual work up procedure was followed. The
crude alcohols were analyzed on a 10 ft. 20% FFAP on Diatoport S column
30
at 150°C with a flow rate of 30 ml of He/minute. Only one peak was
observed under these conditions, but again this is not surprizing since
the steric environment is essentially the same for both isomers. By
preparative glc, these isomeric alcohols were collected together and the
following spectra values were obtained: m.p. 28-29 °C; IR (CDC13 , Cavity
Cell) 3600-3050(s), 2950(s), 2860(s), 1485(s), 1380(s), 1350(m), 1315(s),
1300(m), 1265(m), 1245(m), 1220(s), 1200(s), 1190(m), 1150(m), 1130(s),
1100(s), 1080(w), 1060(w), 945(s); mass spectrum, m/e (rel. intensity)
140(e, 17), 125(40), 122(19), 111(13), 107(31), 97(67), 96(24), 93(62),
85(52), 84(52), 83(24), 82(44), 81(34), 80(16), 79(23), 71(94), 70(28),
69(22), 67(43), 55(51), 43(100), 41(28), 39(22); 1H NMR (CDC13, TMS) 3H
doublet at 0.96 ppm J = 6 Hz, 3H singlet at 1.40 ppm, 10H multiplet at
0.50-2.50 ppm.
The 13C NMR of these alcohols showed 18 lines indicating the
existence of two alcohols. Based on literature values 29 the following
assignments were made: syn-2-endo-methyl-7-methyl-7-norbornanol (KIKa)
13C NMR (CDC13, multiplicity in off-resonance decoupling) C(7), 85.00
ppm (d); C(1), 48.66 ppm (d); C(2), 44.84 ppm (d); C(4), 37.86 ppm (d);
C(3), 31.25 ppm (t); C(6), 28.21 ppm (t); C(5), 21.23 ppm (t); C(9),
19.90 ppm (q); C(8), 17.05 ppm (q). Anti-2-endo-methy1-7-methy1-7-nor-
bornanol (XIXb) 13C NMR (CDC13, multiplicity in off-resonance decoupling)
C(7), 84.27 ppm (d); C(1), 49.02 ppm (d); C(2), 44.92 ppm (d); C( 4 ),
37.13 ppm (d); C(3), 30.58 ppm (t); C(6), 29.06 ppm (t); C(5), 20.63 ppm
(t); C(9), 20.45 ppm (q); C(8), 17.72 ppm (q).
Anal. Calcd for C9 H16 0: C, 77.08; H, 11.50. Found: C,
76.91;
H, 11.44.
The ratio of (XIXa) to (XIXb) was deterMined to be approximately
50:50 by comparing the intensities of the 13C NNR lines associated with
the C(9)s and C(8)s of each compound (see eqs. 5 and 6).
Alkylations of Ketones I, II and III
A 5 ml Erlenmeyer flask equipped with a magnetic stirring bar,
was flash flamed or heated in an oven, cooled and fitted with a rubber
serum cap under nitrogen. To the flask was added the calculated amount
of each ketone in freshly distilled THE or Et 20 along with the internal
standard which was hexadecane. The calculated amount of alkylating
agent was then added to the stirred mixture at the desired temperature.
After the desired tire, usually two hours, the reaction was quenched
with water or saturated ammonium chloride solution. The organic layer was
separated and dried over anhydrous sodium sulfate and then analyzed on
a 10 ft. 20% FFAP on Diatoport S column at 150°C with a flow rate of 30
ml/min. Inverse addition of reactants provided the same results. The
following was the order of elution of the products on this column at
these conditions: ketone (I), 15.3 minutes; ketone (II), 20.2 minutes;
ketone (III), 23.5 minutes; alcohol (XVIIIa), 35.0 minutes; alcohol
(XVII), 39.5 minutes; alcohol XVIIIb), 43.3 minutes; alcohol (XIX),
48,2 minutes and hexadecane appeared after 29.6 minutes. In all cases,
only alcohol XVIIIa was produced when ketone (II) was alkylated. As
noted before, the isomers of alcohol (XIX) could not be separated by any
13 technique but could be identified by C NKR.
31
CHAPTER III
RESULTS AND DISCUSSION
Synthesis of Model Systems for Reduction Studies
The synthesis of 7-norbornanone, (I), was accomplished by a
modified procedure of Gassman and Pape25
(Scheme 1). Hexachlorocyclo-
pentadiene was allowed to react with methanolic potassium hydroxide which
provided 1,1-dimethomytetrachlorocyclopentadiene, (IV). Compound (IV)
was then allowed to react under Diels-Alder conditions with ethylene to
product 1,2,3,4-tetrachloro-7,7-dimethomr.-2-norbornene (V) which was then
dehalogenated with sodium to produce 7,7-dinethoxy-2-norbornene (VI).
7,7-Dimethoxynorbornane (VII) was produced from the hydrogenation of (VI)
with hydrogen in the presence of 5% palladium on carbon catalyst.
Norbornanone (I) was produced from the deketalization of (VII) with 5%
H2 SO4.
Emo-2-methy1-7-norbornanone (II) and*endo-2-methyl-7-norbornanone
(III) were prepared in a' straightforward manner from 7,7-dimethoxy-2-
norbornene (VI) (Scheme 2). Omymercuration of (VI) led to an 80% yield
of pure emo-2-hydroxy-7,7-dimethoxynorbornane (VIII) after distillation.
Chromic acid oxidation31 of the alcohol in pyridine-dichloramethane
afforded 7,7-dimethoxynorbornan-2-one (IX) in an 82% yield following
distillation. This ketone was then converted to the corresponding
methylene compound (X) using methyltriphenylphosphonium bromide and
diasylsodium in dimethylsulfoxide. Catalytic hydrogenation of (X) gave
a 7:1 ratio of (II) to (III) following 5% sulfuric acid catalyzed de-
32
C1
Cl
Me0 OMe
Cl Cl
Cl Cl IV
C l Cl KOH, Me0H
33
IV
C 1
V
Na° ,tBuOH
V .10■111110,
THF, A
VI
H2-Pd/C VI
5% H2SO4
VII
Scheme 1: Preparation of 7-Norbornanone, (I).
Me0 OMe
1) Hg(0Ac) 2 THF•H20 VI
2) NaOH, NaBH >
H 3 CH3
XIa
1) H2 -Pd/c
X 2) Prep
glc OMe
III
1) NAH, DMSO A, 45 min.
IX 2) 03P+CH2 Br
DMSO 3) H2O
X
34
Scheme Preparation of Exo -2-Methyl-7-Norbornanone (II) and Endo-2- Methyl-7-Norbornanone (III).
35
ketalization of (XIa) and (XIb). The ketones and/or ketals were sepa-
rated by gas-liquid chromatography on a 15 ft. 20% carbowax 20 M on
Chramosorb W-NAW column at 135 oC. The NMR showed a chemical shift of
0.96 ppm for (II) and 1.10 ppm. for (III). These values agreed well with
those reported previously. 27
An alternate route for the preparation of (III) was also accom-
plished as shown in Scheme 3. Hexachlorocyclopentadiene was converted
into 5,5-dimethoxy-1,2,3,4-tetrachlorocyclopentadiene (IV) as before.
Propylene diluted with nitrogen was added to (IV) under Diels-Alder
conditions giving 5-methyl-7,7-dimethoxy-1,2,3,4-tetrachloronorbornene,
(XII). (XII) was then dehalogenated in the presence of sodium metal to
give 5-methyl-7,7-dimethoxynorborn-2-ene, (XIII). Hydrogenation of
(XIII) gave (XIa) and (XIb) in a 1:9 ratio. These ketals were then
deketalized to give (II) and (III) in a 1:9 ratio.
The reduction of ketones (I), (II) and (III) was carried out
using LiA1H4 as the reducing agent. For a sumnary of these results, see
Table I. The presence of only one alcohol as the reduced product of
ketone (II) was indicated by glc and 13C NMR. However, it was not
possible to determine whether it was the syn or anti-alcohol. Therefore,
a Birch reduction on ketone (II) was conducted. Since protonation is
faster than equilibration, both the syn- and anti- alcohols should be
produced (eq. 1). It was observed both by glc and IH NER that both the
Me0• OMe Me0 OMe
Cl Cl
CH3CH:=CH2 3 a° , t -BuOH
IV >C1 190-195 °C, 7 HR THF,
36
C l
XII
XIII
XII I
Xlb
Scheme 3: Alternate Synthetic Route to Exo-2-Methy1-7-Norbornanone (II) and Endo-2-Methyl-7-Norbornanone (III).
CH3
Na Liquid NH3
(1 )
II
XVb
XVa
syn- and anti-alcohols were produced in a 20:80 ratio. The alcohols
were separated by glc and found to match the IH NMR spectrum reported in
the literature.32
Under Meerwein-Ponndorf-Verley reduction conditions
using aluminum isopropoxide and isopropanol, only the anti-alcohol from
ketone (II) was produced. This indicates that under equilibrating con-
ditions the anti-alcohol is indeed the most thermodynamically stable.
product. In order to substantiate this, the syn- alcohol was allowed
to equilibrate under Meerwein-Ponndorf conditions employing aluminum
isopropoxide, isopropanol and acetone. The anti-alcohol was formed almost
exclusively except for a trace of the syn-alcohol thus further establish-
ing the anti-alcohol is indeed the thermodynamic isomer.
The 13C NMR spectra of the Birch reduction products were also
obtained. By comparing these spectra with the reduction products of
Lik1H4 with ketone (II), the latter product was confirmed as the syn-
alcohol. Carbon atom-assignments were made by using relative shielding
37
paraneters and off-resonance coupling. It is known that deshielding of
the carbon decreases from tetra-substituted carbons to tri-substituted
to di-substituted with mono-substituted carbons appearing furthest up-
field.
Stereochemistry of 7-Norbornanone Reduction
The reaction of LiAlE'4
with ketone (I) (eq. 2) or ketone (III)
(eq. 5) should produce the corresponding alcohol at twice the rate of
LiA1H4 reduction of ketone (II) to produce the syn-alcohol (eq. 3)
provided "product development control" is not important in this reaction.
If "product development control" is important then, of course, the rate of
attack on ketone (II) to produce the syn-alcohol should be decreased due
to the effect of the exo-2-methyl group on the developing transition
state (product developemnt control).
Whether or not the exo-2-methyl group is sufficiently bulky to
provide a valid test for "product development control" can be evaluated
by comparing the syn-anti-alcohol ratio when LiA1H4 was allowed to react
with ketone (II). If the exo-2-methyl group exerts a significant steric
effect in this systen then significally less anti-alcohol (eq. 4) should
be produced compared to the syn-alcohol in the reaction of ketone (II)
with LiA1H4. In order to test perturbations on the carbonyl group other
than the steric effect exerted by the exo-2-methyl group, the reaction
of LiA1H4 with the endo-2-methyl-7-norbornanone, (III), was also studied.
If only the steric effect of the exo-2-methyl group is significant, then
the reaction of LiA1H4
with ketone (III) to produce the syn- and anti-
alcohol should proceed at the same rate as the reaction of LiA1H4 with
ketone (I) and at twice the rate compared to the formation of the an-
2-exo-methyl alcohol.
38
I
(2)
XIV
( 3 )
Li+
N-
(5) H-ill 0
H/ II
(4)
III CH
3
H
. L
+l H -/ Al—H
Li+
H - 0-Al
(b) H1.1
CH3
H NO
CH3 11
H O
3
SH3 XV Ib
XV Ia
HO
(5)
39
The reduction of'ketones (I), (II) and (III) were carried out
under identical conditions. As noted before, only one reduction product
was obtained for (I) and (II), whereas (III) gave both the syn- and anti-
alcohols according to glc and 13C NMR. By comparing gic and
13C NMR, it
was substantiated that the lone reduction product. of (II) was the syn-
alcohol. Table 1 shows these observations as a result of anti-attack
with respect to the exo-2-methyl group. This shows that the exo-2-methyl
group exerts a significant steric effect with respect to attack at the
7-keto group since no anti-alcohol is observed. When ketones (I) and
(II) were admixed in eqnal molar portions with an insufficient amount of
LiA1H4'
the alcohol products of (I) and (II) were produced in a 2:1
ratio indicating no detectable product development control. Reaction of
(I) and (III) in equal molar portions with an insufficient amount of
LiA1H4 produced the corresponding alcohols in a 1:1 ratio showing that
the endo-2-methyl group has no effect on the rate of reaction of the 7-
keto group. Admixture of ketones (II) and (III) in equal molar ratio
produce the corresponding alcohols in a 1:2 ratio and admixture of
ketones (I), (II) and (III) in equal molar ratio produced the correspond-
ing alcohols in a 2:1:2 ratio when allowed to react with an insufficient
amount of LiA1H4. The data support, the conclusion that anti-attack on
ketone (II) takes place at the same rate as attack from either side of
the carbonyl on ketones (I) and (III) indicating that the exo-2-methyl
group although exerting a significant steric effect (no anti-exo-2-
methyl alcohol formed, eq. 4) does not affect the formation of the syn-
alcohol of ketone (II). When the mole ratio of ketone (II) to (III) was
increased from 1:1 to 2:1 in the presence of an insufficient amount of
41
LiA1H4 the corresponding alcohols produced were in a ratio of 1:1. This
can be explained by the fact that there are now the same number of equal
attack sites on both ketone (II) and (III). When the mole ratio of (II)
to (III) was 4:1, the number of equal attack sites becomes 2:1. On the •
other hand, when the ratio of ketone (II) to (III) was 1:4, the number of
equal attack sites is 1:8 which is what is reflected in the results of
this experiment (Table 1). Futher experiments in TEF and at different
stoichionetric ratios provide additional evidence for the above con-
clusions.
Table 2 compares the Group III& metal hydrides, A1H3, BH3 and
GaH3 reactions with ketones (II) and (III). The results are similar to
those observed for LiA1H4
reduction indicating that the stereochemistry
is independent of the steric requirement of the hydride. Similarly when
LiBH4 , LiA1H4 and LiGaH4 were allowed to react with ketones (I)-(III),
no evidence of "product development control" was observed (Table 3). In
addition when the anion (AIH4 ) was held constant and the cation varied
(Li, Na, and NR), no evidence of "product development control" was
observed (Table 4).
Synthesis of Model Systems for Alkylation Studies
Alkylation of ketones (I), (II) and (III) were carried out using
methylmagnesium bromide in diethyl ether in an attempt to evaluate the
importance of "product development control" when ketones are allowed to
react with org4nometallic alkylating agents. For a summary of these
results see Table 5. Identification of the products of these reactions
is essential just as in the case of the reduction study. The alkylation
of ketone (II) produced only one product as was verified by glc, 1H NMR
and 13C NMR. Assuming that the lone alkylation product was the syn-
alcohol, the anti-alcohol had to be synthesized. A straightforward
method to produce the anti-alcohol was carried out according to Scheme
4. The first step in this sequence was to dehydrate the tertiary alcohol
42
CH CH2
CH CH 3 H2SO4 )
3 mcpB
Scheme 4: Proposed Synthetic Scheme for the Preparation of Exo-2-Methyl-7-Methyl-Anti-7-Norbornanol (XVIIIb).
to the methylene compound folloWed by epoXidation by meta-chloroperben-
zoic acid which is then followed by LiA1H 4 reduction to yield the anti-
alcohol. However, after periodic monitoring by glc, it was noted that a
second peak appeared with a longer retention time than the starting "syn"-
alcohol. This second peak continued to grow until_ it was approximately
1/3 of the starting reactant. This newly formed compound was separated
by glc and identified by 1H .NMR and
13C NMR. By comparing shielding
parameters, as was done for the reduction products identification, this
second compound was identified as the anti-alcohol. The following
sequence is postulated to have taken place (Scheme 5). The first step in
43
CH2
• Ilismsoll•■ 11.••••• AO.C111ESI;;/ 3
CH3
H+ -H 0 CH3 CH3
XVIIIa H+
1[
Scheme 5: Proposed Mechanism for the Acid Isomerization of (XVIIIa) to (XVIIIb).
the process is protonation of the alcohol with loss of water thus forming
the carbonium ion. This can either loose a proton forming the methylene
compound or pick up a hydroxyl group forming either the syn- or anti-
alcohol since the total process is in equilibrium. Evidently the exo-
2-methyl group has a steric requirement regarding the methyl group as
well as the hydroxy group since the anti-alcohol is formed in only 33%
compared to the an-alcohol in equilibrium.
I (6)
XV I I
(7)
(9)
H3
CH 3 XIXb
Stereochemistry of 7-Norbornanone Alkylation
The alkylation of ketone (I) (eq. 6), ketone (II) (eq. 7 and 8)
44
45
and ketone (III) (eq. 9) were carried out under identical conditions. As
noted for the reduction reactions, only one alkylation product was ob-
tained for ketones (I) and (II), Whereas ketone (III) gave both the syn7
and anti-alcohols according to glc, 1H nmr and
I3C nmr. Table 5 shows
the results of the alkylation studies with methylmagnesium bromide.
In Table 6 are recorded the observations of metal alkyl reactions with
ketones (II) and (III). Both tables show essentially the same results
as noted for the reductions studies conducted with the same ketones.
That is, the exo-2-methyl group exerts a significant steric effect with
respect to attack at the 7-keto group since no anti-alcohol is observed.
Also, anti-attack on ketone (II) takes place at the same rate as attack
from either side of the carbonyl on ketone (I) or (III) when in the
presence of an insufficient amount of alkylating agent indicating that
the exo-2-methyl group does not affect the formation of the syn-alcohol
of ketone (II). Therefore it can be concluded that "product development
control" in the alkylation reactions of this model ketone system is not
important compared to "steric approach control".
Reactions of Alkyl Grignard Reagents with Ketone (II)
Since the transition state formed on reaction of CH 3MgBr with exo-
2-methyl-7-norbornanone (ketone II) should not exhibit torsional strain,
compression effects and conformational changes, it is an ideal model
ketone to evaluate "steric approach control" and "product development
control". When CH3MgBr was allowed to react with this ketone, only the
syn-alcohol (O/a) was produced. For this reason, it was recently decided
that exo-2-methyl-7-norbornanone (ketone II) might prove to be a useful
46
model for determining if'a polar or SET mechanism or a combination of
these is responsible for the products obtained from the reactions of
Grignard reagents with ketones.
Due to the large steric effect associated with the exo-2-methyl
group in ketone (II), a polar addition reaction should produce only the
syn-alcohol (eq. 7). If a SET mechanism is in effect, a ketyl would first
be formed, as in the Birch reduction (eq. 1), enabling both the syn- and
anti-alcohols to form in 20:80 ratio when exo-2-methyl-7-norbornanone (II)
was allowed to react with sodium in liquid ammonia. Therefore by allowing
different Grignard reagents to react with ketone (II), observation of the
alkylated anti-alcohol would indicate the possible participation of a
SET mechanism.
Table 7 summarizes the results fran a preliminary study of this
postulation involving the reaction of ethyl, i-propyl, t-butyl, n-hexyl
or i-butyl Grignard reagents with ketone (II). Unfortunately, in no case
was any alkylated product observed. The major product in all cases was
exo-2-methyl-syn-7-norbornanol Ma). Small amounts of exo-2-methyl-
anti-7-norbornanol (XVb) were also observed after quenching with water for
all reactions except for the i-butyl Grignard reagent. The t-butyl
Grignard reagent provided the greatest amount of (XVb) (11%). The other
reagents produced 1-5% of (XVb). The following order of alkyl Grignard
reagents was observed with respect to the formation of (XVb):
t-Bu > i-Pr Et > n-Hex ti i-Bu
9 8H 6 I3H 3 1311 2 1311 1 8H
The reagent which has the most bulky 8-alkyl groups had the least amount
47
of anti-alcohol (XVb) formed. Or in other words, the reagents with the
most I3-hydrogens produced the most anti-alcohol.
From the data, no conculsions can be made concerning polar or SET
mechanisms, but reactions involving benzyl, phenyl, allyl, crotyl, vinyl
etc., which could further our understanding of these mechanisms are now
under further investigation.
CHAPTER IV
CONCLUSION
The concept of "product development control" has been used to
explain the stereochemistry of many reactions in which the observed
isomer ratio reflects the stability of the product. This concept has
been used particularly to explain predominant formation of the most
stable isomer in reactions of LiA1H4
and 110,10r with substituted cyclo-
hexanones. A study of the reaction of LiA1H4 and MeMgBr with 7-norbor-
nanone and its exo-2-methyl and endo-2-methyl derivatives shows that
the most unstable isomer is formed exclusively and hence "product
development control" is not a factor in these reactions. In an attempt
to broaden the scope of this study, three series of reagents were
studied: (1) LiBH4 , LiA1H4 and LiGaH4 , (2) BH3 , A1H3 and GaH3 , and (3)
(CH3 ) 2Be, (CH3 ) 2Zn, (CH3 ) 211g. and (CH3 ) 3A1. In no case was "product
development control" observed. The reactions with the 7-norbornanone
system are similar in nature to those with cyclohexanones, except that
the complicating factors of torsional strain, compression effects and
conformational changes which are present in cyclohexanone systems are
not present in the 7-norbornanone system. The concept of "product
development control" is, therefore, a questionable one in ketone reduc-
tions involving LiA1H4 and alkylations involving MeMOr.
48
Solvent
Ratio Hydride:Ketone
II III I
Recovered Ketone (%)
II III
Et 20 6.00 0.00
Et20 6.00 0.00
Et20 6.00 0.00
Et,0 0.25 0.25 61 80
Et20 0.25 0.25 71 72
Et 20 0.25 0.25 74 59
Et20 0.11 0.11 0.11 69 82 72
till' 6.00 0
THE 0.25 0.25 79 62
Et20 0.22 0.11 169 73
Et20 0.16 0.04 326 82
Et20 0.04 0.16 89 322
95
28
20
21
95
94 94
92 92
14 91
21 92
14 29 88
11 20 91
94 94
15 29 92
21 22 95
31 16 91
4 36 90
Mass Balance . (%)
( % )
Products
Table 1. Reactions of LiA1H4 with Ketones I, II and III in Diethyl Ether and THF. a
a) The hydride was added to 0.032 mmoles ketone at 25 °C for 2 hrs. b) Hydride:Ketone = 6 is equivalent to LiA1H4 :Ketone mole ratio of 1.5:1. c) % of each ketone recovered based on 100% relative to the amount of hydride hydride added. d) % of each product based on 100% relative to the amount of hydride added.
Ratiob
Hydride:Ketone I II III
RecoveredC Ketone (%)
I II III Reducing Agent
Products (%)d
sy...4 off
Mass Balance
Table 2. Reactions of Group IIIb Metal Hydrides with Ketones (II) and (III) in THF. a
6.00 95 95
0.25 0.25 72 16 29 86
6.00 0 96 96
0.25 0.25 71 63 16 30 90
6.00 95 95
0.25 0.25 70 54 18 36 89
a) The hydride was added to 0.032 millimoles ketone at 25 °C for 2 hours. b) Hydride:Ketone = 6 is equivalent to metal hydride:ketone mole ratio of 2:1. c) % of each ketone recovered based on 100% relative to the amount of hydride added. d) %of each:product based on 100% relative to the amount of hydride added.
BH3
BH3
A1H3
A1H3
GaH3
GaH3
Table 3. Reactions of Common Cation Complex Metal Hydrides with Ketones (II) and (III) in THF.a
Reducing Agent
Products (%) d
Ratiob Recoveredc 0H oH
Hydride:Ketone Ketone (%) I II III I II III
off
Mass Balance
LiBH4 6.00 0 97
L iBH 4 0.25 0.25 74 60 16
LiA1H4 6.00 94
LiA1H4 0.25 0.25 79 62 15
LiGaH4 '6.00 0 95
LiGaH4 0.25 0.25 75 59 13
a) The hydride was added to 0.032 millimoles ketone at 25 °C for 2 hours. b) equivalent to complex metal hydride:ketone mole ration 1.5:1. c) % of each ketone recovered based on 100% relative to the amount of hydride added. d) % of each product based on 100% relative to the amount of hydride added.
97
32 91
94
29 92
95
26 86
Hydride:Ketone = 6 is
Table 4. Reactions of Varying Cations of Complex Metal Hydrides with Ketones (II) and (III) in THF. a
ProdUcts (%)d
off
Reducing Agent
Ratiob Hydride:Ketone
I II III
Recoveredc
Ketone (%) I II III
Mass Balance
LiA1H4 6.00 94 94
LiA1H4 0.25 0.25 79 62 15 29 92
NaA1H4 6.00 0 96 96
NaA1H4 0.25 0.25 76 59 15 27 89
NR4AlH4e 6.00 0 95 95
NRA.1H4e 0.25 0.25 76 58 14 26 88
a) The hydride was added to 0.032 millimoles ketone at 25 °C for 2 hours. b) Hydride:Ketone = 6 is equivalent to complex metal hydride:ketone mole ratio 1.5:1. c) % of each ketone recovered based on 100% relative to the amount of hydride added. d) Z of each product based on 100% relative to the amount of hydride added. e) NR4 = tri-n-octyl-n-propylammonium ion and the reagent was prepared in benzene.
Table 5. Methylmagnesium Bromide Reactions with Ketones (I), (II) and (III) in Diethyl Ether and THE.a
Solvent
Ratio
Methyl:Ketone I II III I
Recoveredc Ketone (%)
II III
Products (%) d off oH
Mass
Et2 0 6.00 0 90 90
Et2 0 0.00 0 95 95
Et20 0.00 90 90
Et20 0 0.25 0.25 60 78 28 14 90
Et20 0.25 0.25 70 70 21 21 91
Et20 0.25 0.25 83 64 -- 15 31 96
Et 20 0.11 0.11 0.11 70 81 71 20 11 20 91
Et20 0.22 0.11 173 76 20 19 96
Et20 0.16 0.04 323 82 31 16 90
Et20 0.04 0.16 90 331 4 36 92
THE 6.00 0 -- 93 93
THE 0.25 0.25 81 62 14 29 93
a) The alkylating agent was added to 0.032 millimoles ketone at 25 °C for 1 hour. b) Methyl:Ketone = 6 is equivalent to RMgK:ketone mole ratio of 6:1. c) % of each ketone recovered based on 100% relative
- to the amount of alkylating agent added. d) % of each product based on 100% relative to the amount of alkylating agent added.
Table 6. Reactions of Alkylmetal Reagents with Ketones (II) and (III) in Diethyl Ether. a
Products (%) d
Ratiob Recoveredc Alkylating Methyl:Ketone Ketone (%)
Agent I II III I II III
oN
Mass Balance
Me2Be 6.00 95 95
Me2Be 0.25 0.25 83 64 15 31 97
Me2Mg 6.00 0 95 95
Me2Mg 0.25 0.25 81 60 17 33 96
Me2Zn 6.00 67 67
Me 2Zn - _ 0.25 0.25 - _ 65 43 11 20 70
Me3Al 6.00 65 65
Me3A1 0.25 0.25 60 41 10 20 65
a) The alkylating agent was added to 0.032mmoles ketone at 25 °C for 1 hr. b) Methyl:Ketone = 6 is equivalent to R2M:ketone mole ratio of 3:1. c) % of each ketone recovered based on 100% relative to the amount of alkylating agent added. d) % of each product based on 100% relative to the amount of agent added.
55
Table 7. Reactions of RMgX Compounds With Exo-2-Methyl-7-Norbornanone (II) in Et2a
0 Solvent at Room Temperature for 30 Hours in 2:1 Molar Ratio.
R X Recovered
b
Ketone (%) Yield
bof Syn-
Alcohol(XVa)%(Rel%) c Yieldbof Anti-
Alcohol(XVb)%(Rel%) c
Et Br 35 44(96) 2(4)
i-Pr Br 4 88(95) 5(5)
t-Bu Cl 0 82(89) 10(11)
n-Hex Br 1 90(99) 1(1)
i-Bu Br 2 86 (100) 0(0)
a) The Grignard reagents were prepared by the standard methods. No products other than the reduction products were detected after quenching the reactions with a saturated solution of ammonium chloride. b) Yields were determined by glc and based on internal standards. c) Normalized % syn-alcohol + % anti-alcohol = 100%.
REFERENCES AND NOTES
1. H. O. House, "Modern Synthetic Organic Reactions", W. A. Benjamin, Inc., New York, 1972, p. 45 ff.
2. J. D. Morrison and H. S. Mosher, "Asymmetric Organic Reactions", Prentice-Hall, Inc., Englewood Cliffs, N. J., 1972, p. 116 ff.
3. W. G. Dauben, G. J. Fouken and D. S. Noyce, J. Am. Chem. Soc., 78 ,
2579 (1956).
4. E. Eliel, Y. Senda, J. Klein and E. Dunkelblum, Tetrahedron" Letters, 6127(1968).
5. E. Eliel and R. S. Ro, J. Am. Chem. Soc., 79 5992 (1957).
6. E. Eliel and S. R. Schroeter, J. Am. Chem. Soc., 87, 503 1 (1965).
7. E. Eliel and Y. Senda, Tetrahedron, 26, 2411 (1970).
8. M. Cherest, H. Felkin and N. Prudent, Tetrahedron Lett., 2199 (1968).
9. M. Cherest and H. Felkin, Tetrahedron Lett., 2205 (1968).
10. M. Cherest, H. Felkin and C. Frajernan, Tetrahedron Lett., 379 (1971).
11. M. Cherest and H. Felkin, Tetrahedron Lett., 383 (1971).
12. E. C. Ashby, J. Laemmle and P. Roling, J. Org. Chem., 38, 2526(1973).
13. S. R. Landor and J. P. Regan, J. Chen. Soc., (C), 1159 (1967).
14. J. Klein, Tetrahedron Lett., 4307 (1973).
15. N. T. Anh, O. Eisenstein, J-M Lefour and M-E Tran Hun Dau, J. Am. Chem. Soc., 95, 6146 (1973). C. Liotta, Tetrahedron Lett., 1 (1975);
16. E. C. Ashby and R. D. Schwartz, J. Chem. Educ., 51, 65 (1974).
17. D. F. Shriver, "The Manipulations of Air Sensitive Compounds", McGraw-Hill, New York, 1969.
18. H. Gilman and A. H. Haubein, J. Am. Chem. Soc., 66 1515 (1944).
19. H. Steinbert, "Organoboron Chemistry, Vol. I",Interscience, New York, 1964.
ZU. E. C. Ashby, J. R. Sanders, P. Claudy and R. Schwartz, J. Am. Chem. Soc., 95, 6485 (1973).
56 :
57
21. A. E. Finholt, A. C. Bond and H. I. Schlesinger; J,;'A . Chem. Soc., 69, 1199 (1947).
22. G. W. C. Milner, Analyst, 80, 77 (1955).
23. J. S. Newcomer and E. T. McBee, J. Am. Chem. Soc., 71 946 (1949).
24. P. G. Gassman and J. L. Marshall, Organic Synthesis, V, p. 424.
25. P. G. Gassman and P. G. Pape, J. Org. Chem., 29, 160 (1964).
26. R. Greenwald, M. Chaykovsky and E. J. Corey, J. Org. Chem., 28, 1128 (1963).
27. D. A. Lightner and D.• E. Jackman, J. Am. Chem. Soc., 96, 1938 (1974).
28. S. Winstein and E. T. Stafford, J. Am. Chem. Soc., 79, 505 (1957).
29. N. K. Wilson and J. B. Stothers, Top. Stereochem., 8, 1 (1974). G. Levy, "Topics in Carbon-I3 NMR Spectroscopy, Vol. I and II", Wiley and Sons, New York, 1974.
30. R. K. Bly and R. S. Bly, J. Org. Chem., 28, 3165 (1963).
31. G. I. Poos, G. E. Arth, R. E. Beyler and L. H. Sarett, J. Am. Chem. Soc., 75, 422 (1953).
32. T. L. Gertelsen and D. C. Kleinfelter, J. Ore . Chen., 36, 3255 (1971).
59
CHAPTER
INTRODUCTION
Background
Considerable interest in organic synthesis at present is centered
in the use of transition metal hydrides for the hydrometallation of al-
kenes and alkynes. Stoichiometric amounts of transition metal hydrides
have been reported to reduce effectively unsaturated organic compounds.
Conjugated C=C or C=N-bonds1-6
have been reduced and organic halides7
have been reductively dehalogenated by [HFe(CO) 4 ] and by several
derivatives of "Cull".8-11
In protic media12
the same transformations
can be accomplished by [HFe3 (C0) 11 ]: Wailes and Schwartz have reported
independently that hydrozirconation of alkenes13-15
and alkynes
also involve a hydrometallation intermediate.
The hydrozirconation of alkenes was shown to proceed through the
placement of the zirconium moiety at the sterically least hindered
position of the alkene. The authors argued that the formation of the
product involved either the regiospecific addition of Zr-H to a terminal
alkene or Zr-H to an internal alkene followed by, rapid rearrangement via
Zr-H elimination and readdition to place the metal again in the least
hindered position.
Transition metal hydrides are also used as catalysts for reactions
of unsaturated hydrocarbons such as hydroformylation, hydrogenation,
hydrosilation and isomerization.18
Recently, the reduction of alkenes
19,20 and alkynes by the reagent LiAlH
4-transition metal halide was reported.
60
Although one might assume that this reaction proceeds through a hydro-
metallation intermediate, deuterolysis of the reaction mixture shows that
only titanium compounds are effective in the formation of the hydrometal-
lation intermediate. Other first row transition metal compounds (e.g.
NiC12 and CoCl2 ) are effective in catalyzing the formation of reduction
products although no evidence for a stable transition metal intermediate
has been found.
Purpose
This research has centered around an investigation of the hydro-
metallation of alkenes and alkynes using less expensive and more readily
available catalyst systems than has been used so far. The importance of
forming the hydrometallated intermediate rather than the reduction product
(alkane or alkene) lies in the formation of an organometallic compound
that can be easily functionalized. Although hydroboration proceeds
readily between an olefin and diborane in THE in the absence of a cat-
alyst, the 0-B bond in relatively stable and not as susceptible to func-
tionalization as are C-Mg or C-Al compounds. Unfortunately, MgH2 and
A1H3 do not hydrometallate alkenes or alkynes at all readily compared
to B2 H6' • however, reaction does take place when certain transition metal
halide catalysts are present.
CHAPTER II
EKPERIMENTAL SECTION
Apparatus
Reactions were performed under nitrogen or argon at the bench
using Schlenk tube techniques or in a glove box equipped with a recircula-
ting system using manganese oxide columns to remove oxygen and dry ice-
acetone traps to remove solvent vapors.22
Calibrated syringes equipped
with stainless steel needles were used for transfer of reagents. Glass-
ware and syringes were flamed or heated in an oven and cooled under a flow
of nitrogen or argon. All inorganic and organic compounds including
internal standard solutions were prepared by weighing the reagent in a
tared volumetric flask and diluting with the appropriate solvent.
All melting points are corrected and all boiling points are un-
corrected. Proton NMR spectra were determined at 60 MHz with a Varian,
Model A-60 or Model T-60 or at 100 MHz with a JOEL Fourier Transform
spectrometer, Model PFT-100. The chemical shift values are expressed in
ppm (6 values) relative to a Me4Si internal standard. The mass spectra
were obtained with a Hitachi (Perkin-Elmer), Model RMU-7 or a Varian,
Model M-66, mass spectrometer, GLPC analyses were carried out on a F and
M Model 700 or Model 720 gas chromatograph. The it spectra were determined
with a Perkin-Elmer, Model 621 or Model 257, infrared recording spectro-
photometer. High pressure reactions were carried out in an autoclave
rated to 15,000 psi obtained from the Superpressure Division of American
Instrument Company of Silver Springs, Maryland.
61
62
Analytical
Gas analyses were carried out by hydrolyzing samples with 0.1 M
hydrochloric acid on a standard vacuum line equipped with a Toepler pump.23
Aluminum was determined by adding excess standard EDTA solution to hydro-
lyzed samples and then back titrating with standard zinc acetate solution
at pH 4 using dithizone as an indicator. Lithium reagents were analyzed
by the standard Gilman double titration method (titration of total base
then titration of total base after reaction with benzyl chloride). 24
The amount of active C-Li was determined by titrating the active reagent
with dry 2-butanol in xylene using 2,2'-diquinoline as an indicator.
Amine was analyzed by injecting hydrolyzed samples with an internal
standard on the gas chromatograph. Carbon, hydrogen analyses were carried
out by Atlanta Microlab, Inc., Atlanta, Georgia.
Analysis of all products arising from the quenching of reactions
of alkenes and alkynes with hydride reagents with H20, D20, CO, CO2 , 12 ,
02 or carbonyl compounds were identified by glc and/or nmr and isolated
by gic techniques and compared to authentic samples obtained commercially
or synthesized by proven methods. All nmr spectra were obtained in CDC1 3
or benzene-d6 using Me
4Si as the internal standard.
Materials Solvents
Fisher reagent grade anhydrous diethyl ether was stored over
sodium, then distilled under nitrogen from LiA1H4 and/or sodium-benzo-
phenone ketyl.
Fisher reagent grade tetrahydrofuran (THF) was dried over NaA1H 4
and distilled under nitrogen using diphenylmethane as a drying indicator.
63
Fisher reagent grade benzene and hexane were stirred over concen-
trated H2SO4' washed with Na2 CO3, then distilled water, dried over an-
hydrous MgSO4 and distilled from NaA1H4 under nitrogen or argon.
Alkenes
1-Octene (b.p. 122-123 °C), 1-methyl-l-cyclohexene (b.p. 110-111 °C),
styrene (b.p. 145-146 °C), cis-2-hexene (b.p. 67-68 °C), trans-2-hexene
(b.p. 68-69 °C), 1-hexene (b.p. 63-64 °C), methylenecyclohexane (b.p. 102-
103 °C), 2-ethyl-l-hexene (b.p. 119-120°C), cyclohexene (b.p. 82-83°C),
neohexene (b.p. 40-41 °C), 2,3-dimethyl-2-butene (b.p. 72-73 °C), and 2-
methy1-2-butene (b.p. 35-38°C) were obtained from Chemical Samples
Company or Aldrich Chemical Company and distilled and stored over 4 A
molecular sieves.
Alkynes
1-Hexyne (b.p. 70-71 °C), 2-hexyne (b.p. 83-84 °C), 4-octyne (b.p.
132-133 °C), 1-phenyl-1-propyne (b.p. 185-186°C), diphenylethyne (b.p.
170°C, 19 mm), and 1-octyne (b.p. 124-125 °C) were obtained from Chemical
Samples Company or Aldrich Chemical Company and distilled and stored
over 4 A molecular sieves.
1-Trimethylsily-l-octyne. Into a 250 ml three-necked round bottom
flask fitted with a Teflon coated magnetic stirring bar, rubber serum cap
and a pressure equalizing addition funnel fitted with a three-way stop-
cock connected to an argon filled manifold equipped with a mineral oil
filled bubbler was placed 14.8 ml (11.0 g, 100 moles) of 1-octyne in
50 ml of distilled hexane. To this stirred mixture, cooled with an ice-
water bath, was slowly added 50 ml of 2.03 M n-butyllithium in hexane
via syringe. After the addition, 30 ml of distilled hexane was added and
64
then the reaction mixture was allowed to warm to room temperature and
stirred for three hours. The resulting white slurry was again cooled by
an ice-water bath, and 13 ml (11.1 g, 112 moles) of trimethylchloro-
silane (distilled from quinoline) was slowly added over a period of 1/2
hour. The resulting mixture was allowed to warm to room temperature
and afterwards, stirred for three hours. Then 50 ml of water was added,
slowly at first. The hexane layer was separated and washed twice with
water, dried over anhydrous sodium sulfate and the solvent removed vsing
a rotary evaporator. The crude product was distilled to give 15.1 g
(0.083 mole, 83% yield) of 1-trimethylsilyl-1-octyne, b.p. 45-49 °C (1.0-
1.2 mm) (lit. 25 b.p. 49-50°C, 1.0 mm); IR(neat, film) 2950(s), 2860(s),
2180(s), 1470(m), 1250(s), 850(broad-s), 770(s), 710(m); NMR (CDC13, TMS)
9H singlet at 0.14 ppm, 3H multiplet at 0.68-1.06 ppm, 8H multiplet at
1.08-1.66 ppm, 2H multiplet at 2.00-2.38 ppm; mass spectrum, m/e (rel.
intensity) 182(M+, 0.04), 168(18), 167(100), 154(3), 139(2), 123(3),
109(12), 96(13), 83(12), 73(26), 59(14).
Anal. Calcd for C11H22Si: C, 72.44; H, 12.16. Found: C, 72.23;
H, 12.13.
Ketones and Aldehydes
Fisher Certified A.C.S. grade acetone was dried over MgSO 4, then
filtered, distilled from P205
and stored over 4 A molecular sieves.
Finton 4-t-butylcyclohexanone was sublimed under nitrogen.
Eastman benzophenone and Aldrich benzaldehyde were distilled under
vacuum and stored under argon in the dark.
Preparation of Inorganic Reagents
Lithium aluminum hydride (Alfa Inorganics) solutions in THE or
diethyl ether and sodium aluminum hydride (Alfa Inorganics) solutions in
65
THE were prepared by ref luxing LiA1H 4 and NaA1H4 in the appropriate
solvent for at least 24 hours followed by filtration through a fritted
glass filter funnel in the dry box. The clear solutions were standardized
for aluminum content by EDTA and by hydrogen using standard vacuum line
techniques. 20
Alane, A1H3 , was prepared by the reaction of 100% H 2 SO4 with LiA1H4
in THE at -78 °C and filtered in the dry box26 using a fritted glass filter
funnel and dry Celite as a filter aid.
H2A1C1, HA1C12 , H2A1Br, HAlBr2 and H2A1I were prepared in THE by
the redistribution reactions of A111 3 and A1C13 , A1Br3 or A113 . 27 These
reagents were characterized by analyzing for aluminum and hydrogen.
H2A10But , HA1(0But ) 2 , H2A1age, HA1(0Me) 2 , H2A1NP4, H2A1NEt2 ,
H2A1N(SiKe3 ) 2 , HAl(NPr2) 2 , HAl(NEt 2 ) 2 and HAl[N(Sitle3 ) 2 ] 2 were prepared
by simply adding the appropriate alcohol or amine to A1H 3 in THE in a 1:1
or 1:2 molar ratio. Hydrogen was involved during the addition and the
reaction was complete within 15 to 20 minutes except in the case of the
reaction involving diisopropylamine or 1,1,1,3,3,3 -hexamethyldisilazane
in which case three hours of reaction time were required. The HA1(2
compounds were identified by their Al-H stretching frequency assignments:28
-1 - i 1 HA1(0But ) 2 , 1850 cal ; HA1(0Me)2, 1
1840 cm ; HAl(NPr 2 ) 2 , 1810 cm ;
-1 • HAl(NEt 2 ) 2 , 1820 cm ; HAl[N(SiMe3 ) 2 ] 2 , 1800 cm
-1. These compounds were
also analyzed for their aluminum content by titration with EDTA and back
titration with zinc acetate and also by hydrogen analysis using standard
vacuum line techniques.
These alkoxy and amino reagents were also prepared in benzene by
removing the THE from the appropriate reagent under vacuum followed by
addition of freshly distilled benzene. This procedure was repeated
three times until all TEF had been removed. The solutions were then
analyzed for aluminum and hydrogen content.
Activated Lill was prepared by the hydrogenation of t-butyllithium
or n-butyllithium at 4000 psi for 12 hours at room temperature in hexane.
The resulting LiH slurry was removed via syringe under an argon atmo-
sphere or in a dry box.
Sodium hydride as a 50% oil dispersion was obtained from Alfa
Inorganics. The oil was removed by repeated washing and decantations
using freshly distilled hexane.
Lithium and sodium trimethylaluminohydrides were prepared by the
equal molar addition of a benzene, diethyl ether or THE solution of tri-
methylaluminum (obtained from Ethyl Corporation and distilled under
vacuum in a dry box) to a lithium or sodium hydride slurry in the
appropriate solvent. The addition was carried out in an appropriate
sized one-necked round bottom flask equipped with a magnetic stirring
bar and a pressure equalizing addition funnel while being cooled with an
ice-water bath. The additon funnel was fitted with a rubber serum cap
which was attached to an argon filled manifold connected to a mineral oil
filled bubbler by a syringe needle. After the addition and stirring
(usually 10 minutes) the reaction mixutre became a clear, pale brown
solution which as analyzed for aluminum by EDTA. titration and lithium
and sodium by flame photometry.
Sodium bis(2-methoxyethoxy) aluminum hydride (Vitride T) was ob-
tained as a 70% toluene solution from Matheson, Coleman and Bell.
Lithium and sodium bis-diethylaminoaluminohydride were prepared
66
either by adding with stirring at 0°C the stoichiometric amount of
diethylamine to a THE solution of lithium or sodium aluminum hydride or
by adding the stoichiometric amount of bis-diethylaminoalane (preparation
discussed above) to activated lithium or sodium hydride. The clear, pale
yellow-brown THE solutions were analyzed for lithium, aluminum and
hydrogen by the standard methods described in the Analytical Section. If
benzene solutions were desired, the THE was removed under vacuum and re-
placed by freshly distilled benzene. This procedure was repeated three
times. The amount of THE according to glc which remained was < 5%.
Bis-(cyclopentadieny1)-dinethyl Titanium
Into a 250 ml one-necked round bottom flask, equipped with a
magnetic stirring bar and a pressure equalizing addition funnel fitted
with a three-way stopcock with one arm connected to an argon filled
manifold which in turn was connected to a mineral-oil filled bubbler,
was placed 7.47 g ( 20.0 mmoles) of bis-(cyclopentadienyl)titanium
dichloride, Cp 2TiC12, and 100 ml of freshly distilled hexane. To this
stirred and cold (-78 °C) mixture was added 50 ml of 1.35 M MeLi in n-
hexane (67.5 mmole) slowly over a 2 hour period. Then the mixture was
allowed to warm to room temperature and stir for an additional two hours
followed by slow addition of 2 ml of freshly distilled methanol. The
organic layer was removed via syringe and the precipitate was washed
repeatedly with distilled n-hexane. The hexane fractions were combined
and stored in a freezer at -20°C. The solvent was removed from a 10 ml
aliquot by flash evaporation. The resulting yellow-orange crystals
melted at 96-97 °C (lit. 29 m.p. 97 °C). The nmr spectra in benzene and
'EMS showed 5 H singlet at 5.71 ppm and 3 H singlet at 0.053 ppm.
67
Dicobalt Octacarbonyl
Because of the toxic nature of dicobalt octacarbonyl and cobalt
tetracarbonyl hydride, all procedures30, 31
were conducted in a well
ventilated hood.
Into a 500 ml stainless steel autoclave were placed 150 ml of
hexane and 15 g of cobalt (II) carbonate (0.126 mole) obtained from Alfa
Inorganics. The autoclave was flushed three times with carbon monoxide
and then filled with an approximately equimolar mixture of hydrogen and
carbon monoxide to a pressure of 3500 psi. The autoclave was heated with
agitation to 150°C and these conditions were maintained for three hours.
The autoclave was then cooled to room temperature, the gases vented
slowly and the autoclave opened in a well ventilated hood.
The clear dark solution of dicobalt octacarbonyl was pipetted
from the autoclave via syringe and filtered through a fritted glass
filter. The filtrate was then stored over night in a freezer (-20 °C)
during which time large, purple crystals were formed. The solvent was
decanted and the crystals were dried using a stream of dry carbon
monoxide. The dry crystals (16.5 g, 78%) melted with some decomposition
at 50-51 °C (lit.30 m.p. 51-52°C). The product was stored under a carbon
monoxide atmosphere.
Cobalt Tetracarbonyl Hydride (Cobalt Hydrocarbonyl)
A 500 ml three-necked round bottom flask was equipped with a
pressure-equalizing addition funnel fitted with .a rubber serum cap, an
inlet tube which was constricted at the tip to about 2 mm and inserted
nearly to the bottom of the flask, and a three-way stopcock with one arm
attached to a 36 mm X 110 mm drying tube containing a mixture of P205
68
69
and glass beads. This apparatus in turn was attached to a cold trap
immersed in liquid nitrogen and with its outlet tube connected to a
mineral oil filled bubbler.
Three grams of cobalt octacarbonyl (0.009 mole) were placed into
a 300 ml ground glass Erlenmeyer flask. Twenty ml of freshly distilled
pyridine (0.248 mole) was added and the flask attached to a mercury
filled bubbler to prevent entry of air and to monitor the formation of
carbon monoxide which was complete in a couple of minutes. The resulting
solution was added to the addition funnel. A solution of 25 ml of conc.
H2 SO4 and 75 ml of water was cooled to 0-5°C and added to the round-
bottom flask cooled by an ice-water bath. Before the pyridine solution
was added, the apparatus was purged with carbon monoxide at a flow of
about 300 cc/min. The slow addition was performed over 1 hour. The
flask was then purged with carbon monoxide for another 15 minutes which
allowed the cobalt hydrocarbonyl to be transferred from the reaction
flask to the cold trap. After the carbon monoxide was allowed to evaporate
slowly leaving behind the HCo(C0) 4 , about 2 g (95%) of product was isolated
in this manner. Hexane solutions were prepared by the addition of freshly
distilled hexane and stored at -78 °C over sodium sulfate. Analysis of
the above solutions was by the method of Orchin, et a1. 32 which in-
corporated the titration of the cobalt hydrocarbonyl by tris-(1,10-phen-
athroline) nickel (II) chloride. The analysis was also carried out
in the following manner: Into a 50 ml Erlenmeyer flask is placed a
Teflon stirring bar, 10 ml of water and several drops of ethanolic
phenolphthalein. The flask is fitted with a rubber serum cap and then
purged with carbon monoxide through syringe needles. A 1 ml - sample of
the hexane solution is added. The resulting two-phase system is then
stirred continuously during titration with 0.01 N sodium hydroxide.
Metal Salts
AlC13' A1Br3
and AlI3 (Alfa Inorganics) were sublimed just prior
to use and crushed to a fine powder in the dry box.
The following transition metal halides were obtained from Fisher:
TiC13 , TiC14 , CrC1 3 , MnC1 2 and ZnBr . Alfa Inorganics supplied the
following transition metal halides: VC13, FeC13, FeCl2, CoC12, NiC12,
Cp2TiC12 , Cp2ZrC12 , Cp2Ni(PPh3)C1, Ni(acac) 2 , Ni(PPh3 ) 2Br2 , allyl-
Ni(dpe)Br, Cp 2VC1 2 and the polymer bound benzyl titanocene dichloride
(Grubb's catalyst). All transition metal halides were opened only in the
dry box and used without further purification.
General Reactions of Alkenes and Alkynes
A 10 X 18 mm test tube with a Teflon coated magnetic stirring bar
was flamed and cooled under a flow of argon or nitrogen. Transition metal
halide (5 mole % in most cases) was transferred to a tared test tube in
the dry box. The flask was sealed with a rubber serum cap, removed from
the dry box, reweighed and connected by means of a syringe needle to a
nitrogen or argon-filled manifold equipped with a mineral oil-filled
bubbler. However, in the case of Cp 2TiC12 , a saturated solution in
benzene (0.02 M) or THF (0.125 M) was prepared; however, the solutions
had to be made fresh each day. One or two ml of THF or benzene was
introduced into the reaction vessel and then the olefin or alkyne added.
The mixture was stirred for 2 to 3 minutes before the reagent was added.
The the reaction mixture was stirred at room temperature or at higher.
temperatures, depending upon the reactants, for up to 40 hours in some
70
71
cases. In general the reaction between the terminal alkenes and internal
alkynes were complete in 20 minutes. The reactions were quenched by
various means (see below-General Quenching Techniques) and worked up by the
regular method (addition of water, extraction with THF, diethyl ether
or hexane and drying over MgSO 4 ). Most products were separated by glc
using a 6 ft. 10% Apiezon L 60-80 column with a helium flow rate of 45
ml/min: 1-octene (110°C, oven temperature), 1-methyl-l-cyclohexene (50°C),
2-ethyl-l-hexene (50°C), cyclohexene (50°C); a 20 ft. 10% TCEP column
with a helium flow rate of 45 ml/min for 1-hexene, cis-2-hexene, trans-
2-hexene, neohexene, 2-methyl-2-butene and 2,3-dimethy1-2-butene (45°C,
flow rate 25 ml/min); 2-hexyne, 1-octyne and 1-hexyne (70°C, flow rate
45 ml/min); 1-trimethylsilyl-1-octyne (100 oC, flow rate 45 ml/min); 1-
phenylpropyne (125 °C, flow rate 45 ml/min); a 10 ft. 5 % carbowax 20 M
column for diphenylethyne (200 °C, flow rate 60 ml/min). The yield was
calculated by using a suitable hydrocarbon internal standard for each
case (n-C12H26 , n-C14H30 or n-C16H34) and the products were identified
by comparing the retention times of authentic samples with the products
under similar conditions and/or by coinjection of products and authentic
samples obtained commercially or synthesized by proven methods.
Yields of cis-stilbene ( 6 6.60, vinyl H), trans-stilbene ( 6 7.10,
vinyl H) and 1,2-diphenylethane ( 6 2.92 benzyl H) were determined by NMR
integration and based on total phenyl protons. However, the ratio of
cis-stilbene to trans-stilbene was also checked by glpc.
General Quenching Techniques
Quenching with H20. After the desired reaction time for the
catalytic hydrometallation reaction described above, the reaction was
72
quenched with water or a saturated solution of ammonium chloride to
produce the protonated species. The amount of recovered starting
material (alkene or alkyne) and products were determined by the methods
described above.
Quenching with D 20. The same procedure used for quenching with
H2O was followed. The amount of recovered starting material and products
were determined by glc methods and conditions described above. Each
product was collected from the gas chromatograph and submitted for mass
spectrum analyses. The corrected percent of deuterium incorporation for
the product was calculated by comparing the protonated species' mass
spectrum with the deuterated species' mass spectrum and by subtracting
the contributions of naturally occurring isotopic components from each
molecular ion peak. This procedure was followed for all unsaturated
substrates observed under these reaction conditions.
The regioselectivity of reactions was monitored by use of NMR for
styrene, 1-phenyl-1-propyne and 1-trimethylsilyl-1-octyne. In the case
of styrene, 98% of the product, ethylbenzene, was deuterated with 90%
of the deuterium located on the a carbon as indicated by a quartet at
2.64 ppm (2H) J = 8 Hz and a doublet (3H) at 1.23 ppm, J = 8 Hz.
For 1-phenyl-1-propyne, 95% of the product, cis-l-phenylpropene,
was deuterated with 90% of the deuterium located on the number 1 carbon
as indicated by a quartet of triplets at 5.78 ppm, J = 7 Hz (1H) and a
doublet at 1.89 ppm, J = 7 Hz (3H). The other isomer showed a singlet at
1.89 ppm and a multiplet at 6.36-6.54 ppm. It was identified as trans-
1-phenylpropene.
The products from the reduction of 1-trimethylsilyl-1-octyne were
prepared independently. Cis-1-trimethylsilyl-1-octene was prepared from
the hydrogenation of 1-trimethylsily1-1-octyne with 5% Pd/C used as the
catalyst and 95% ethanol used as the solvent and monitored until the de-
sired amount of hydrogen was absorbed. The cis-isomer was collected and
purified via glpc under the aforementioned conditions. The trans-isomer
was also detected by glpc. The cis-isomer when coinjected under glc con-
ditions with the product of the hydroalumination reaction, showed a trace
characteristic of only the cis-isomer. The trans-isomer obtained from the
hydrogenation reaction had an identical retention time as the minor pro-
duct from the hydroalumination reaction.
The following data were obtained for cis-1-trimethylsily1-1-octene:
IR (neat, film) 2960(s), 2940(s), 2860(m), 1600(m), 1470(m), 1260(s), 850
(broad-s); NMR (CC14' TMS) 9H singlet at 0.14 ppm, 11H multiplet at 0.74-
2.66 ppm, 2H quartet at 2.15 ppm, J = 8 Hz, 1H doublet at 5.49 ppm, J =
13.0 Hz, 1H doublet of triplets at 6.33 ppm, J = 14 Hz and 7 Hz; mass
spectrum, m/e (rel. intensity) 184(M+, 2), 170(13), 169(70), 141(4), 125(4),
114(23), 109(13), 99(26), *5(15),
Anal. Calcd for C11H24Si:
73(100),
C, 71.65;
67(9), 59(91), 44(21),
H, 13.12. Found: C,
41(14).
71.54;
H, 13.14.
The nmr spectrum of the trans-isomer matched the spectrum found
in the literature:33
NMR (CC14 , TMS) 9H singlet at 0.16 ppm, 11H
multiplet at 0.6-1.6ppm, 2H multiplet at 2.1 ppm, 1H doublet at 5.6 ppm,
J = 18 Hz, 1H doublet of triplets at 6.0 ppm, J = 18 H z
and 6 H.
When 1-trimethylsily1-1-octyne was allowed to react under the
catalyzed hydrometallation conditions described above, 87% of the product,
cis-1-trimethylsilyl-1-octene, was deuterated and approximately 90% of
73
74
the deuterium was located on carbon number 2,as indicated by the following:
NMR (CDC13, TMS) 9H singlet at 0.14 ppm, 11H multiplet at 0.80-1.56 ppm,
2H broad singlet at 2.14 ppm, 1H singlet at 5.49 ppm. The other isomer
( trans- 1-trimethylsilyl-1-octene) had no deuterium incorporation but
showed a triplet at 6.33 ppm which corresponded to the expected splitting
pattern of the non-deuterated trans-isomer ,
Quenching with I 2. A known concentration of iodine in benzene
was prepared and a stoichiometric amount was added to the catalytic
hydroalumination reaction after the desired time. This mixture was then
allowed to stir at room temperature for 1 hour. Afterwards, water was
added followed by a saturated sodium thiosulfate solution. The organic
layer was then separated and dried over sodium sulfate and analyzed by
glc or NMR. This procedure was followed for the following substrates:
1-octene, 1-hexene, cis-2-hexene, trans-2-hexene, 3-hexene and 2-hexyne.
The iodoalkane-compounds prepared in this manner were confirmed by com-
parison of authentic samples purchased from Eastman Chemical Company or
by quenching the corresponding alkyl Grignard with iodine and worked up
in the aforementioned manner. The iodoalkenes were identified by NMR
and discussed below.
When the catalytic hydroalumination reactions of internal alkynes
were quenched with D 20, the products are the cis-alkenes which were
confirmed by coinjection of authentic samples on the gas chromatograph.
From Zweifel s 34 work, it was reported that hydroaluminated compounds
quenched with iodine maintain their regiochemistry. Therefore the
iodoalkene obtained from the quenching of the 2-hexynyl reaction with
iodine produced 2- and 3-iodo-cis-2-hexene. NMR was used to distinguish
75
. /4 between the two isomers: R/C=C\ ( CH3
singlet at 2.36 ppm) and CH3
(CH3 doublet at 1.63 ppm J = 7 H e). The following spectra R CH3 were also obtained:IR(neat, film) 2940(s), 2920(s), 2850(m), 1625(m),
1460(s), 1430(s), 1380(s), 1120(s), 1070(m), 1055(m), 1040(m); mass
spectrum, m/e (rel. intensity) 210(M+, 100), 181(10), 168(3), 128(4),
127(6), 83(19), 67(6), 55(91), 41(3).
Anal. Calcd for C6H11I: C, 34.30; H, 5.28. Found: C, 34.51;
H, 5.25.
Quenching With Carbonyl Compounds. After the desired catalytic
hydroalumination reaction had taken place, a stoichiometric amount of
desired carbonyl compound (acetone and benzaldehyde) was added, followed
by additional stirring for 10 hours accompanied by maintaining the
desired temperature with an oil bath. The reaction mixtures were then
worked up by addition of water. The organic layer was then separated,
dried over MgSO4 , filtered and analyzed by NMR techniques.
One of the products arising from the quenching of 1-octene and 4-
octyne reactions with acetone, the reduction product, isopropanol, was
determined by observation of the methyl group attached to the carbinol
carbon (doublet at 1.2 ppm). Acetone was determined by observation of
the methyl groups attached to the carbonyl carbon (2.05 ppm). 1-Octene
was identified by the vinyl proton multiplet at 4.8-5.2 ppm (2H) and
5.52-6.24 ppm (1H). 4-Octyne was identified by the triplet corresponding
to the methyl group at 0.98 ppm. 1-Octene and 4-Octyne were also deter-
mined by glc analyses. The addition products were determined by nmr ob-
servation of the methyl groups which appearred as singlets at 1.22 ppm
and by comparing these spectra with those obtained from the reaction of
76
corresponding octyl Grignard reagents with acetone. The integrity of the
octenyl addition compound was established by the methyl singlet at 1.22
ppm and the appearance of a multiplet at 5.5-6.3 ppm.
For the products arising from the quenching of 1-octene and 4-
octyne reactions with benzaldehyde, benzaldehyde was determined by the
singlet associated with the aldehydic proton at 9.94 ppm, the reduction
product, benzyl alcohol, was determined by the singlet found at 4.58 ppm,
the addition product from 1-octene was determined by the methine proton
triplet at 4.60 ppm corresponding to the Grignard addition product's
nmr spectrum and the addition product from 4-octyne was determined by
the appearance of the vinyl multiplet at 5.5-6.3 ppm and also by glc,
mass spectral and it analyses.
When HAl(NPr2 ) 2 was allowed to react with 1-octene or 4-octyne
in the presence of 2 mole % Cp 2TiC12 and then allowed to react with
benzaldehyde, the major product was diisopropylbenzylamine in both
cases. The following data were obtained: nZ5D 1.5300; N1R (CDC13, TMS)
12H doublet at 1.01 ppm, J= 5 H z ; 2H multiplet at 2.66-3.34 ppm; 2H
singlet at 3.63 ppm; 5H multiplet at 7.10-7.40 ppm; mass spectrum, pile
(rel. intensity) 191(1+, 8), 176(56), 134(5), 132(3), 114(2), 106(6),
105(6), 91(100), 84(2), 77(6), 65(6), 51(3), 43(5), 42(4), 41(5), 39(3);
IR (neat, film) 3090(w), 3070(w), 3030(m), 2970(s), 2940(m), 2880(w),
1605(m), 1495(m), 1470(m), 1455(m), 1385(s), 1375(s), 1210(s), 1180(s),
1155(m), 1140(m), 1120(m), 1075(w), 1050(w), 1025(s).
Anal. Calcd for C 13 H21N: C, 81.61; H, 11.07. Found: C, 81.49;
H, 11.00.
77
When HAl(NEt 2 ) 2 was allowed to react under similar conditions, the
major product was diethylbenzylamine: n 25D 1.5325; NMR (CDC1 3 , TMS) 6H
triplet at 1.07 ppm, J=6 H z ; 4H quartet at 2.42 ppm, J=6 H z ; 2H singlet
at 3.64 ppm, 5H multiplet at 7.10-7.40 ppm; mass spectrum, m/e (rel.
intensity) 163(M+ , 6) 148(19), 133(6), 132(7), 118(7), 109(11), 108(25),
107(19), 106(79), 105(83), 91(68), 86(7), 80(11), 79(21), 78(22), 77(100),
65(8), 58(18), 51(42), 45(18), 43(26), 39(14); IR (neat, film) 3090(w),
3070(w), 3030(w), 2970(s), 2940(m), 2880(w), 1605(m), 1495(m), 1470(m),
1450(m), 1380(s), 1215(s), 1185(s), 1145(m), 1120(s), 1070(m), 1030(s),
950(s).
Anal. Calcd for C11E.17N: C, 80.92; H, 10.50. Found: C, 81.05;
H, 10.45.
General Reactions of Complex Metal Hydrides
The same general procedure used for the bis-dialkylaminoalane
reactions of alkenes and alkynes were followed as described above. The
complex metal hydrides used were lithium and sodium aluminum hydride,
sodium bis-(2-methoxyethoxy) aluminum hydride and lithium and sodium tri-
methylaluminum hydride. The reactions were quenched with H 20, saturated
solution of ammonium chloride, D2 () or a benzene solution of iodine and
worked up as described above. The products were identified by the above
procedures. General Reactions of LiH or NaH
The same procedure for the catalytic hydroalumination reactions
of alkenes and alkynes with bis-dialkylaminoalanes described above were
also followed for this part of the investigation. The identification
of the products proceeded as before by glc analyses and by comparison
with authentic samples.
78
General Reactions for the Carbonylation of Simple and Mixed Metal Hydrides
The simple and complex metal hydrides were carbonylated at 4000
psi using a high pressure apparatus at room temperature. Diethyl ether,
THE or hexane was used as the solvent and the reactions were carried out
over a 24 hour period. Gas-liquid chromatography was used to confirm
the production of methanol, ethanol and/or methyl formate by comparison
with authentic samples.
•
CHAPTER III
RESULTS AND DISCUSSION
Reactions of Alkenes
The monosubstituted olefin, 1-octene, was chosen as a represent-
ative olefin for the initial evaluation of catalysts in the study of
hydroalumination of olefins. Bis-diisopropylaminoalane, [H.A1(NP4)21 ,
was chosen as a representative alane and was allowed to react. with
1-octene in the presence of 5 mole percent of.various transition metal
compounds (eq. 1). Nearly quantitative yields of octane were obtained
PhH HAl(NP4) 2 + 1-Octene + Catalyst ), Octane (1)
RT, 1 hr
when TiC13 , TiC14 , CoC12 , NiC12 , Cp 2TiC1 2 , polymer bound benzyltitanocene
dichloride, CpNi(Ph3P)C1, Ni(PEt3 ) 2Br2 and allyl-Ni(dep)Br were used as
catalysts (Table 8). The intermediate formation of the octyl aluminum
compound was monitored by the percent deuterium incorporation observed
when the reaction was quenched with D 20. Even though the yields of octane
were high, only the titanium catalysts provided high yields of products
showing deuterium incorporation. The best catalyst for this reaction
(eq. 1) appears to be titanocene dichloride, Cp 2TiC12 .
Scheme 6 outlines the proposed mechanism involved in the catalytic
hydrometallation process. The transition metal halide is proposed to
react with the disubstituted alane to produce the disubstituted chloro-
aluminum compound and the transition metal hydride. The hydroalumination
species is presumed to be the transition metal hydride and the reducing
7 9
L = Halogen, Cp
MLn H
Ln M - H M = Transition Metal X = OR, NR2, CI
80
MLnCI + HAIX2 MLn H CIAIX2
A
L i4 -H ■
3 +
I
H 0
ALKANE
Scheme 6: Proposed Catalytic Hydrometallation Mechanism.
81
ability of the reagent is believed to be due to d-orbital overlap between
the metal ion and the unsaturated carbon-carbon bond. Under this assump-
I 10 II ,10 tion, Cu (d ) and Zn (a ) have no empty d-orbitals to overlap with the
olefin, and MnII
(d5 ) with the d-orbitals half filled would be predicted '
to have a lower activating ability. Under the conditions described above,
Cu, Zn and Mn catalysts are considerably slower reacting than the titanium
cobalt or nickel catalysts. This explanation is consistant with the
results obtained.
After the addition of all the reactants and reagents, a blue-
violet color appeared which represented the formation of the transition
metal hydride.35
As stated above, Cp2TiC1
2 was shown to be an effective
•catalyst;36.
however, when Cp2TiMe
2 was employed as the catalyst, no
color change was observed and none of the anticipated products was formed
and the starting material was recovered. This can be explained by the
fact that the chloride in titanocene dichloride is a better leaving group
than the methyl group in dimethyltitanocene and therefore Cp 2TiC12 is
more acceptable as a catalyst involving simple displacement by the bis-
dialkylaminoalanes than Cp2TiMe2 .37
The transition metal hydride first
coordinates with the unsaturated substrate forming complex A.
The next step involves f3-hydride addition of the transition metal
hydride to the coordinated unsaturated substrate forming compound B.
Hoffman38
proposed that this equilibrium depends an the donation of the
electron density on the metal thus stabilizing the olefin complex.
IV Therefore by using a d 0 metal (§,g, Cp2T1 C12 ), specific activation of
the alkene or alkyne should be accomplished; in other words, the d0
metal-alkyl complex B should predominate over the unsaturated complex A
at equilibrium. Once the alkyl transition metal compound is formed, a
transmetallation step with the disubstituted alane takes place. Hart and
Schwartz 39 have shown that alkyl groups can be transferred from zirconium
to aluminum simply by treatment with A1C13'
It is suggested that a
similar process occurs in the analogous titanium system. The transition
metal hydride complex which is also formed in this step can now react
with more substrate to continue the hydrometallation process.
In order to determine the nature of the reaction intermediate
when alkene is allowed to react with HAl(NR 2 )2 in the presence of
transition metal halides, deuterium incorporation experiments were
carried out by quenching the reaction mixtures with deuterium oxide.
The products were collected by preparative glc and the deuterium content
(%) was measured by the molecular ion peak ratio of deuteriated and
non-deuterated product in the mass spectrum. These results are also
listed in Table 8. The only experiments giving high yields of product
and deuterium incorporation occurred with the titanium catalysts (TiC1 3 ,
65%; TiC14 , 80% and Cp 2TiC12 , 932). A notable exception was the polymer
bound benzyltitanocene dichloride (Grubb's Catalyst). The catalyst
produced 99% of the reduced alkene but no deuterium was incorporated
when the reaction was quenched with D 20. It was hoped that with a polymer
bound catalyst, the hydrometallated species could easily be removed from
the catalyst thereby reducing the probability of further side reactions
taking place. These results imply that the hydrometallated transition
metal halide intermediate is not stable under the conditions studied
except for the titanium catalysts. In other words, the transmetallation
reaction involving alkyl transfer from the transition metal to aluminum
82
proceeds only for the reactions with titanium catalysts (eq. 2).
/ HA1(NR2 ) 2,4 + 'Till' (2)
H i H Al(NR2 )
Several experiments were carried out in an attempt to stabilize
the carbon-transition metal bond by varying the ligands attached to the
transition metal. It is expected that ligands are capable of stabilizing
the transition metal compounds by dispersing the d-orbitals of the
transition metal through the attached ligands. The addition of triphenyl-
or triethylphosphine to nickel chloride did not have much of an effect
on the percent deuterium incorporation. However, the addition of cyclo-
pentadienyl ligands to TiCl4 had pronounced effect on deuterium incor-
poration (93%).
It was also shown that the amount of deuterium incorporation in-
creased when the solvent was changed from THF (78%) to benzene (88%)
and when the atmosphere was changed from nitrogen (88%) to argon (93%)
(Table 9). The changing of a nitrogen atmosphere to one of argon
increased the amount of deuterium incorporation observed because evident-
ly N2 was "fixed" to the titanium atom in a side reaction. The changing
of solvents from THF to benzene also increased the amount of deuterium
incorporation presumably due to the inability of benzene to donate a
hydrogen atom during homolytic cleavage of the R-Ti compound as would be
the case for THF.
Table I0 contains the observed results from reaction of a series of
alkenes with HAl(NPr2) 2 in the presence of 5 mole percent Cp 2TiC12 . All
83
84
of the reactions were conducted in benzene at 60 °C for 12 hours except
that for the 1-octene experiment which was over in 10 minutes at room
temperature. Relative rates for these reactions were found to be styrene
I-octene > 3,3-dimethy1-1-butene > methylene cyclohexane > 2-ethyl-l-
hexene >> cis-2-hexene trans-2-hevma>> cyclohexene >>> 2-methy1-2-
butene ild2,3-dimethyl-2-butene (1, 1methylcycIohexene. These rates
parallel the rates found by Schwartz for the hydrozirconation reactions.15
The percent deuterium incorporation,which monitored the production
of the intermediate alkyl aluminum compounds, was high for styrene, 1-
octene, cis-2-hexene, trans-2-hexene, methylene cyclohexane and 2-ethyl-
1-hexene (98 93, 83, 81, 72 and 75% respectively). The notable exception
was 3,3-dimethyl-l-butene which gave a high yield of 2,2-dimethylbutane,
reduction product, in this case 2,2-dimethylbutane, was obtained (99%),
but only IO% deuterium incorporation of the product was observed. Cyclo-
hexene and the trisubstituted or t:etrasubstituted olefins, 1-methylcyclo-
hexene, 2-methyl-2-butene or 2,3-iimethyl-2-butene, provided only small or
no amounts of the corresponding alkanes. This implies that the greater
the steric bulk of the olefin the slower the reaction.
According to Scheme 6, the critical step in the catalytic hydro-
metallation reaction is the production of A which leads to the formation
of B in a regioselective manner where the transition metal proceeds to
the terminal carbon for primarily steric reasons. This result is sub-
stantiated by the fact that only 1-iodooctane is produced when the
catalytic hydrometallation reaction product of 1-octene is quenched with
iodine (Table 11). When 2-ethyl-l-hexene is the substrate, the transition
metal hydride complex can arrange itself in a manner where the compression
85
effect is minimal (Figure 1). This allows the normal reaction to proceed
Figure 1
and hence a large percentage of deuterium incorporation. But for the
neohexene reaction, the now present methyl group exerts a larger
compression effect than the hydrogen in the 2-ethyl-1-hexene system.
This, therefore, prevents the proper alignment of the groups and forces
the reaction to follow a different pathway and consequently, no deuterium
incorporation is observed in the product.
It should also be noted that 98% of ethylbenzene produced from
the styrene reaction was incorporated with deuterium after quenching
with D20. Of that 98%, 90% of the deuterium was located in the 1
position, PhCH(D)CH3 . As will be discussed latter, the titanium
hydridochloride catalyst could complex with the 7 cloud of the phenyl
ring (Scheme 7, complex A 1 ) permitting the titanium to be close to the
carbon holding the phenyl ring. This phenomenon will once again be
evident for the 1-phenylpropyne reaction.
86
When 1-hexene, 1-octene, cis-2-hexene and trans-2-hexene are
allowed to react under conditions listed above, but quenched with iodine,
only the corresponding I-iodo-compounds were produced (Table 11). This
observation is consistent with previous findings of Schwartz observed for
the hydrozirconation process.15 It has been shown that zirconium (IV)
and titanium (IV) salts can catalyze the isomerization of secondary
aluminum reagents40 to produce prinary aluminum reagents. These iso-
merizations may occur through.reversibly f3-hydride elimination and re-
addition mechanisms mentioned before in reference to Scheme 6.
Reactions of Dienes
When dienes (Table 12) were allowed to react under these conditions,
the conjugate dienes (1,3-butadiene and 1,3-hexadiene) yielded only .a
mixture of butenes and hexenes without any alkanes being observed. When
quenched with D20 only 41% of the product showed deuterium incorporation.
The non-conjugated diene, 1,5-hexadiene, in TIC' yielded a mixture of
10% 1-hexene and 80% hexane with 75% deuterium incorporation when quenched
with D 20. However, in benzene only, methylcyclopentane was observed with
90% deuterium incorporation when quenched with D 20. Evidently, in
benzene cyclization takes place (eq. 3)41
whereas in THE cyclization
87
R2N
R2Ic
cat benzene
(3)
A/*[ >
,7-,4
.AlA NR2
NR2
S = solvent
does not take place because THE is strongly coordinated to aluminum and
therefore is difficult to be displaced since the electron demand of the
aluminum is somewhat satisfied. In benzene, however, rearrangement is
still possible.
Survey of Catalysts of Hydrometallation of Internal Alkenes
According to the literature, all attempted catalytic hydrometal-
lation reactions involving internal olefins result in at least some iso-
merization to the terminally substituted products after quenching. It
would be desirable to be able to hydrometallate internal alkenes without
isomerization. Therefore a series of transition metal halides were
allowed to react as catalysts with HAl(NPr2 ) 2 and cis-2-hexene in benzene
at 60°C for 24 hours (Table 13). After the allotted time, the reactions
were quenched with a benzene solution of iodine. The products were an-
alyzed by glc. Under these conditions no 2-iodohexane was observed al-
88
though the cobalt and nickel catalysts produced nearly quantitative yields
of hexane (99%). The titanium catalysts (TiC13 , TiC14 and Cp 2TiC12 ) pro-
vided 1-iodohexane in 55, 56, and 75% yields respectively accompanied by
5-13% trans-2-hexene. When the reaction with Cp 2TiC1 2 was allowed to
proceed at room temperature for 24 hours in benzene or THE, only a trace
of 1-iodohexane, none of the 2-iodohexane, 85% cis-2-hexene, and 14%
trans-2-hexene were detected by glc. Therefore one can conclude from
these data that the catalytic hydrometallation of internal alkenes pro-
ceeds with isomerization.
Reactions of Alkynes
The results obtained from the reactions of various alkynes with
HAl(NPr2
)2 and 5 mole percent Cp
2TiC12 in benzene at room temperature
under an argon atmosphere are tabulated in Table 14. In the case of the
internal alkyne (2-hexyne, 2-octyne, 1-phenylpropyne and diphenylethyne,
Table 14) reactions, a small amount (5, 10, 3 and 0% respectively) of the
alkane was observed by glc. This by-product could be minimized to about
3% conducting the reaction at 0 °C in an 80:20 mixture of benzene/THF for
2 hours. The major products were cis-2-hexene (94%), cis-2-octene (90%),
cis-1-phenyl-1-propene (96%) and cis-stilbene (96%) with only 1-4% of the
trans-isomer observed. All the products were confirmed by comparing the
it spectra of identical samples obtained by the hydrogenation of the
starting materials (Pd/C and hydrogen) or samples obtained from the Aldrich
Chemical Company. When the reaction mixtures were quenched with D 20, 96-
97% of the cis-alkenes (Table 14) contained deuterium according to mass
spectroscopy. When the reaction with 2-hexyne was quenched with iodine,
the nmr spectrum showed a 53:47 ratio of 2-iodo-cis-2-hexene to 3-iodo-
89
cis-2-hexene (Table 14). This result is expected, on steric as well as
electronic grounds since there is little difference in either effect be-
tween alkynyl methyl and propyl groups.
A study was conducted in order to determine (1) the effect of
temperature on the products of reaction and (2) the product composition
from 0 to 100% reaction. A representative internal alkyne, 2-hexyne,
was allowed to react with HAl(NPr 2)2 in the presence of various amounts
of Cp2TiC12 . The results are summarized in Tables 15 and 16. When the
reaction was carried out at room temperature for 16 hours only 52%
deuterium incorporation in the product was observed; however, when the
reaction was carried out at 0 °C, 91% deuterium incorporation was ob-
served, once again indicating formation of the intermediate cis-2-
hexenylaluminum compound in high yield. This reaction is believed to be
reversible as shown in Scheme 6; however, this result indicates that
either the alkenyltitanium or alkenylaluminum compound is not stable at
room temperature under hydrometallating conditions and subsequently
undergoes decomposition to produce H2 which, in the presence of catalyst,
will hydrogenate the alkyne to the alkene which in turn can be hydro-
genated to the alkane. This suggestion would account for the low
deuterium content of the product following work up procedures. A
separate experiment showed that hydrogen was indeed produced when
HAl(NPr2)2 was added to Cp 2TiCl
2 at room temperature. Hydrogen was
also produced when diisobutyloctenylaluminum in benzene was added to
5 mole percent of Cp 2TiCl2 . The diisobutyloctenylaluminum was prepared
according to the procedure of Wilke which allowed 1-octyne to react
with diisobutylaluminum hydride at 60 °C. This result also supports
the above proposition. The amount: of alkane which was produced from
I) cat + 2 H2 2) D2)
2
2 I) cat + 2 H2 2) D2)
ilst 3 H2 2) D20
4
S Hexyl-CIT-H + HAl(NR 2 ) 2 Cat ) Hexy1-41=rAl(NR2 ) 2 + H2
2
90
2
•
Hal (NR2 ) 2 cat Hexyl-C11=iC[Al(NR
3 .2 ] 2
cat 3
•
HA1(NR2 ) 2 > Hexyl-CH2C[Al(NR2 ) 2 ] 3
4
Hexyl-CH2CH3
Hexyl-CH2 CH2D
Hexyl-CH2CHD2
Hexyl-CH2 CD3
2 ) Hexyl-COIC-D
Scheme 8: Proposed Mechanism for the Polydeuterated Octane Formation.
catalyst had little effect on the product ratios.
In order to circumvent the problem of producing a wide variety of
products discussed above, the trimethylsilyl compound was prepared
according to equation 4. When 1-trimethylsilyl-1-octyne was allowed to
the reaction of 2-hexyne with HAl(NPr2)2 in the presence of Cp2
TiC12
was reasonably constant throughout the reaction at 0°C indicating that
the alkane is formed in the initial stages of the reaction by hydro-
genation. Also considerable trans-olefin is formed when the temperature
is increased from 0 to 25 °C indicating the advantage of lower reaction
temperatures. If the alkyne was added after the catalyst and alane were
allowed to react with stirring for a few minutes under a slight vacuum,
the amount of alkane produced decreased to less than 3%, but the amount
of alkene also decreased to 45% with 94% deuterium incorporation.
As the amount of catalyst was increased (Table 17) the rate also
increased with no loss in deuterium incorporation. A reasonable cat-
alyst concentration is considered to be 5 mole %.
If 1-phenylpropyne was allowed to react under similar conditions
(Table 14), 96% of cis-I.-phenyl-I.-propene, 3% of phenylpropane and 1% of
trans-1-phenyl-1-propene were produced upon quenching with water. The
cis-alkene produced in this manner matches the nmr, ir, mass spectrum
and refractive index of cis-I.-phenyl-I.-propene obtained from the catalytic
hydrogenation of 1-phenylpropyne. When the catalytic hydrometallation
reaction was quenched with D20, the product contained 97% deuterium
according to mass spectroscopy. The nmr spectrum showed a 10:90 ratio
of 2-deuterio-cis-1-phenyl-1-propene to 1-deuterio-cis-l-phenyl-1-propene
(Table 14). However, Eisch42 reported that quenching the reaction of di-
isobutylaluminum hydride and 1-phenylpropyne at 50°C with D20 produced
an 80:20 ratio of the 2-deuterio-cis-l-phenyl-1-propene to 1-deuterio-
cis-I-phenyl-I-propene indicating that attack occurred at the least
hindered carbon 4 to 1. The reverse regiochemistry observed for the
90
Cp.\ AI
Cp/ !NI H
C C HI
Cp
71. Hi *1' Cp
7:3›-C C C if3
111 A2
Scheme 7: Proposed Mechanism For The Catalytic Hydrometallation Of 1-Phenylpropyne.
91
93
hydrometallation reaction products reported herein indicates, according
to Scheme 7, that the formation of the titanocene hydridochloride-alkyne
complex determines the regiochemistry of the products. Complex A l is
less hindered than complex A 2 , because the cyclopentadienyl and phenyl
groups in complex Al can adapt a staggered arrangement, but the cyclo-
pentadienyl and methyl groups in complex A2 cannot. Therefore complex
Al would be favored over complex A 2. Of course, the titanium hydrido-
chloride catalyst could also be complexed to the w cloud of the phenyl
ring thus positioning the titanium closer to the carbon holding the phenyl
ring. Following the 8,-hydride addition, the transmetallation step with
HAl(NPr)2 would generally have the same environment and therefore would 2
have little effect on the overall reaction products.
When a representative terminal alkyne (e.g. 1-octyne) was allowed
to react under the above conditions, an approximately 50:50 mixture of
alkane to alkene was observed. When the reaction mixture was quenched
with D20, the alkene showed a relatively high percentage of deuterium
incorporation (78%); however, the octane showed deuterium only to the
extent of 55%. According to mass spectroscopy, four octanes [octane-d0
(45%), octane-d 1 (22%), octane-d, (18%) and octane-d
3 (15%)] were pro-
duced which Scheme 8 can account for.42 The deuterium incorporation
exhibited in the octane products indicates that metallation or depro-
tonation of the acetylenic hydrogen occurs with subsequent addition of
H-Al across the multiple bond.42
Also, with hydrogen present along with
a catalyst, hydrogenation reactions are possible and thereby accounting
for the large number of products. The lowering or raising the tempera-
ture, increasing the reagent/alkyne ratio or increasing the amount of
94
CH3 (CH2 CEE-H + n-BuLi hexane
CH (m2) 5c=r—Li + Butane1
(4)
I. CH3(CH
2 ) 50a-Li + C1SiMe3 hexane
„I CH3(CH
2)5C=C--SiMe
3 + LiC14,
react under the usual catalytic hydrometallation conditions (Table 14),
only 5 percent of the totally saturated compound, octane-d 0 , and the
trans-alkene were produced with the cis-alkene being the predominant
product (90%). This product's nmr, it and mass spectra matched those
obtained for the product from the hydrogenation of 1 -trimethylsily1- 1-
octyne by Pd/C and hydrogen. When the hydrometallation reaction product
was quenched with D 20, 80% of the product contained deuterium according
to mass spectroscopy. NMR analysis of the products showed that approx-
imately 90% (Table 14) of the deuteriated compound was the cis-1-tri-
methylsily1-2-deutereo-1-octene. This result is somewhat surprizing in
that one would expect the aluminum to be adjacent to the more electro-
negative silicon atom. 42 However, as shown earlier in Scheme 7, the
regiochemistry of the product is determined by formation of the inter-
mediate alkynyl titanium compound. The bulky trimethylsilyl group
hinders the approach of the titanium catalyst and therefore 90% of the
newly formed titanium compound is located in the 2-position. An in-
dication of this is that a similar result was obtained when HAl(NEt 2 ) 2
was used as the reagent.
95
Further Reactions With Carbonyl Compounds and Oxygen
In order to evaluate the extent to which further reactions (other
than reactions with D20 or halogens) can be applied to the newly formed
hydroaluminated species, stoichiometric amounts of acetone, benzaldehyde
and benzophenone were added to the reaction mixtures and allowed to react
at room temperature and at 30 °C for an additional 24 hours. The results
are listed in Table 18. First, HAl(NPr2) 2 was allowed to react with
acetone, benzaldehyde, 4-t-butylcyclohexanone and benzophenone in a 2:1
ratio in benzene. The alane reduced the non-enolizable carbonyl com-
pounds , benzaldehyde and benzophenone, in 100% yield to the correspond-
ing alcohols. However, acetone was reduced to yield 40% of isopropanol
with 35% of the starting ketone recovered and the remaining ketone
presumably lost through condensation resulting from enolization. This
alane also reduced 4-t-butylcyclohexanone to provide 18% axial alcohol
and 82% equatorial alcohol in a 45% overall yield with 30% recovered
ketone. The relative ratio of alcohols is approximately the same as
A1H343
in THF (19% axial alcohol). However, in benzene, HAl(NPr 2)2 in a
2:1 ratio of reagent to ketone provided 30% of the axial alcohol and 70%
of the equatorial alcohol in a 45% yield with 30% of the ketone recovered.
To a lesser degree, this result is reminiscent of the trialkylaluminum
reactions in benzene compared to those in THF.
When 1-octene was added to HAl(NPr)2 and allowed to react under 2
hydrometallation conditions and then the reaction mixture added to
acetone or vice versa, there was produced only 5% isopropanol, 2% of the
addition product (2-methyl-2-decanol), and 70% acetone. The analogous
reaction with 4-octyne. produced essentially the same results. When
96
benzaldehyde was allowed to react with the hydrometallated species formed
from the reaction of 1-octene or 4-octyne with HAl(NPr2) 2 and Cp 2TiC12 ,
the major product was benzyldiisopropylamine (90%). When HAl(NEt 2 ) 2 was
used as the hydrometallating agent,90% of benzyldiethylamine was produced.
When benzophenone was allowed to react with HA1(NEt2
)2 or HAl(NP4) 2 ,
62% of the corresponding amine (Ph 2CHNR2 ) was produced in addition to 38%
benzhydrol. Only a trace of the expected addition product was observed.
All products were determined by methods discussed in the experimental
section.
These latter developments were only recently observed in this
study. Therefore more experiments must be carried out in order to
determine the extent to which this reaction can be of synthetic utility.
Stoichiometry, solvent, temperature and rate studies must be carried out
in order to maximize the production of the amine. Also, other enolizable
and non-enolizable carbonyl compounds should be allowed to react under
these conditions along with the necessary control experiments. That is,
independently prepared alkyl and alkenyl bis-dialkylaminoaluminum com-
pounds should be allowed to react under these conditions with and without
the presence of catalyst. If a catalyst is necessary then a survey of
catalysts should be carried out.
It is proposed that (3-reduction by the newly formed RAl(NR 2 ) 2
compound or direct reduction by HA1(NR2 ) 2 of the carbonyl group takes
place first forming an alcoholate which can then be displaced intra-
molecularly (a) or intermolecularly (b) (eq. 5). If this is the case,
HAl(NR2 ) 2
Ph
° NR R) (H)
R2t a NR
(5)
2
H A
RA1(NR2)2 )
'Al(NR2 ) 2
- olef i
Ph
97
P\ i/NR2
then alcohols should be allowed to react under these conditions which
would indicate if the alcoholate is indeed the intermediate. If alcohols
can produce tertiary amines under these mild conditions, it would repre-
sent a major development since there are very few methods to prepare
amines from alcohols in one step without the use of exotic reagents and/or
drastic conditions..45
98
It is known that R3A1 compounds undergo rapid oxidation with 0 2
to produce, on hydrolysis, alcohols 19 or with CO2
to produce tertiary
alcohols or carboxylic acids depending on the reaction conditions.46
When oxygen or carbon dioxide was passed through the catalytic hydro-
metallation reaction mixture of :L-octene and HA1(NPr2)2 in the presence
of Cp 2TiCl2 , only octane was observed by glc after work up with saturated
ammonium chloride or 10% HC1. The reason for this may be argued on in-
ductive electron withdrawing grounds. On the electronegativity scale,47
carbon has a value of 2.5 while the value of nitrogen is 3.0. This means
that nitrogen withdraws electron density from the adjacent carbon atom
thereby strenthening the aluminum-carbon bond compared to R3A1 compounds
(Figure 2). The electron donating ability of the NR2 group by resonance
stronger bond NR2 less reactive
than AIR3
Figure 2
is diminished because the orbital size difference between aluminum and
nitrogen prohibit good overlap thus the overall effect is inductive
explaining the lack of reaction of RA1(NR2)2 with CO2
or 02' Credence
is given this supposition by the fact that when triethylaluminum in
benzene was oxidized in the presence of 5 mole percent of Cp TiC1 2 2'
n-propanol was prepared.
99
Survey of Substituted Alanes
As stated above, bis-dialkylaminoalanes in the presence of 5 mole
percent Cp 2TiC12 proved to be an excellent reagent for accomplishing
catalytic hydrometallation. However, in order to study the scope of
reagents successful in this reaction, it was decided to investigate the
effects of other substituted alanes on the catalytic hydroalunination
of a representative olefin (e.g. 1-octene). Table 18 lists the results
of this investigation. As discussed above the bis-diisopropyl- and bis-
diethylaminoalanes produced high yields of I-iodooctane (80 and 86%,
respectively) and octane-d / (93 and 90%, respectively) when quenched with
the appropriate reagent. The mono- and dichloroalanes under catalytic
hydroalumination conditions produced 97% octane with 85% and 83% deuterium
incorporation, respectively. Sato, et al., 19 reported a 90% yield of
hexane when A1H2C1 or A1HC1
2 was allowed to react with 1-hexene in the
presence of 2% TiC14 at 15
oC for 6 or 16 hours, respectively. Unfortun-
ately, these reactions were not quenched with D 20, halogen, etc. processes
which would indicate the extent to which the hydroaluminated species was
formed. The decrease in rate for HA1C12 compared to H2A1C1 observed by
Sato, et al.1,9c
was also observed in the present work. We found that 2-5%
of Cp 2TiC12 increased the rate of reaction compared to that observed by
Sato, I9b,c but a relative rate decrease was also observed going from
H2A1C1 to HA1C12' 2 and 6 hours, respectively. The relative decrease in
reaction rate maybe due to the electron withdrawing ability of the
chlorine atom which results in the strengthening of the Al-H bond (Figure
3) thus making it less reactive than A1H3 towards Cp 2TiC12 as well as the
H-----Al
much weaker bond more reactive than HA1(NR2
)2
SiMe3
S
R
H Cl fiER
.7 ) \ 1--4C H Al -7
1 H Al
Nless reactive stronger bond\ weaker bond H /f Cl
than A1H3
less reactive more reactive ,f than H2A1C1
than HA1C12
11.
Figure 3
transmetallation step described in Scheme 6. When two chlorine atoms are
present, the effect is even greater resulting in even a slower rate. This
effect is even noticeable when comparing HAl(NR 2) 2 compounds with HA1C1 2 .
The electronegativities of nitrogen and chlorine are equal (3.047
);
however, because alkyl groups are attached to the nitrogen atom, the
effect is diminished somewhat by inductive electron donation, and there-
fore the H-Al bond is weaker and accordingly, more reactive, and thus
these reactions are complete in approximately 15 minutes. In the case of
his- 1,1,1,3,3,3-hexamethyldisilazoalane, this effect is also observed.
When this reaction was quenched with iodine, both the 1- and 2-iodooctane
were produced in a relative ratio of 53:47 with a 90% yield , whereas, when
all of the other reactions were quenched with a benzene solution of iodine,
only the 1-iodooctane was obtained. This latter result maybe explained
"100
101
in the following manner. On the electronegative scale aluminum, silicon
and nitrogen have values of 1.5, 1.8 and 3.0, respectively.47
Silicon is
a very good electron donor and the nitrogen is a good electron acceptor.
Therefore, the amount of electron density on the aluminum is not de-
creased as much as in the other cases discussed above. This results in a
more reactive H-Al bond. In Scheme 9, we proposed that there exists an
equilibrium between the two complexes of the titanium hydride and the
olefin (A1 and A
2). If the formation of the titanium hydride or the
transmetallation step depends on the reactivity of the alane, then it can
be seen that the more reactive the alane the greater the possibility of
forming compound B 2 which would eventually lead to the formation of 2-
iodooctane. This, evidently, was the case with the disilazane derivative
reaction which was over in about 10 minutes at room temperature. Once
compounds Bland B2 are formed, both are likely to undergo a transmetal-
lation reaction with HAl(NR2 ) 2
to form C1 and C2. These products in turn
can produce the 1- and 2-deuteric of iodoalkenes when quenched with D 20 or
iodine.
When a more electronegative atom such as oxygen (electronegative
value of 3.547
) is incorporated into the reagent, the reaction is slowed
down even more because of a stronger and less reactive H-Al bond which was
demonstrated for the H2A1ONe, HA1(ale) 2 , HA1(0Pr i ) 2 and HA1(OBut ) 2
reagents. For the alkoxy derivatives which were insoluble in THE or
benzene only the monomethoxy reagent produced 10% octane after 24 hours
at room temperature and quenched with water. The diisopropoxy and di-t-
butoxy derivatives which were more soluble in THE produced only a 15%
yield of octane.
Cp, LTi
H \
Cp2TiC12 + HA1(NR2 ) 2
/1
C1A1(NR2 ) 2 + Cp 2TI\
102
Bl
HAl(NR2 ) 2 111 HAl(NR2 ) 2 1[
1(NR2 ) 2 A. (m2 ) 2
H R H H
1 Cp2Tk
\H
Cl ./
Cp Ti 2 \ H
Scheme 9: Proposed Mechanism for the Production of the Kinetic and
ThermodynaMicyroducts from the-Catalytic HydrometallatiOn.
Reaction
103
Hydrometallation With LiH
Recently Caubere and co-workers showed that NaH-RONa-MXt systems
were capable of reducing halides,48,49
ketones, alkenes and alkynes.50,51
However, these reactions were not catalytic, but did provide 75-95% of
the corresponding alkanes, alcohols, alkanes and alkenes respectively.
In our continuing studies of catalytic hydrometallation reactions we
investigated the reactions of activated LiH with alkyl halides, carbonyl
compounds, alkenes and alkynes in the presence of catalysts. Table 19
lists the results of the reaction of activated LiH and transition metal
halides with 4-t-butylcyclohexanone in 1:1:1 ratio. Without the transi-
tion metal halide only a 5% yield of the alcohols were obtained with 55%
being the equatorial alcohol. However, an equal molar ratio of VC1 3 with
respect to LiH provided an 86% yield of the alcohols with 82% being the
axial isomer, FeC13 reduced 68% of the ketone providing 74% of the axial
alcohol. A. catalytic amount of VC13
(5 mole percent) was added but under
these conditions only 3% of the alcohols were obtained with basically the
same stereoselectivity observed for the stoichiometric reaction.
Table 20 represents the reactions of LiH-VC1 3 in 1:1 ratio with
other carbonyl substrates. The ketones (4-t-butylcyclohexanone, 3,3,5-
trimethyIcyclohexanone, 2-methylcyclohexanone and camphor) were reduced in
high yields with the axial or exo alcohols predominant (78, 90, 92 and
95% respectively). The aldehydes (hexanal and benzaldehyde) and esters
(ethyl benzoate and ethyl n-butyrate) were also reduced to their respective
alcohols (95, 97, 93 and 95%) in high yields. In the case of ester re-
ductions, small amounts of the corresponding aldehydes were also observed.
When alkenes were allowed to react under similar condit ions, only
the terminal alkenes (1-6ctene, 2-ethyl-l-hexene and methylene cyclo-
hexane) reacted to form the corresponding alkenes in high yield (Table
21). However, none of the reactions produces high percentages (29 and
30%) of deuterium incorporation when quenched with D 20. Changing sol-
vents from THE to benzene had no effect on the outcome.
In an attempt to improve on the work reported by Caubere, we
decided to explore, in a cursory sense, the possibility of catalytic
(5 mole %) hydraaetallation using LiH or NaH. Table 22 summarizes the
results of this study. Of all the transition metal halides surveyed,
Cp2TiC12 produced the most encouraging results when allowed to react with
activated LiH and a representative olefin, 1-octene. A 77% yield of
octane was produced, and if the reaction was quenched with D 20, 50%
deuterium incorporation was observed by mass spectroscopy. It was some-
what disappointing that Nail did not produce yields larger than 5% of
octane, because from an economic point of view, Nail is inexpensive and
easy to handle.
If alkynes (Table 23) or enones (Table 24) were allowed to react
with LiH-VC13
in 1:3 ratio, only starting material was recovered in all
cases after 36 hours at 45 °C (except for the reaction of cinnaldehyde
which produced 90% of the 1,2 reduction product). Evidently the other
enones either enolized or were too bulky for reaction to take place.
Hydrometallation Reactions With Main Group Complex Metal Hydrides
Sato, et a1.19 reported that titanium tetrachloride and zirconium
tetrachloride catalyzed the addition of lithium aluminum hydride to
olefinic double bonds to produce the organoaluminate. They reported,
for example, that 1-hexene produced 99% n-hexane when allowed to react in
104
105
THE at room temperature for 30 minutes in the presence of 2 mole % of
TiC14.
However, the reactions were not quenched with D 20 a process
which would monitor the production of the hydrometallated intermediate.
When these reactions were quenched with bromine, only 70% of the corre-
sponding bromide was obtained in most cases. Therefore the product ob-
tained from the quenching of the reaction with H 2O was misleading.
In order to clear up this question, we investigated the above
reaction in more detail. Lithium aluminum hydride, sodium aluminum
hydride, lithium trimethylaluminum hydride, sodium trimethylaluminum
hydride, lithium and sodium bis-dialkylaminoaluminum hydride and Vitride
(sodium bis(2-methoxyethoxy)aluninum hydride) were allowed to react with
a series of alkenes and alkynes in a ratio of alkene or alkyne to reagent
of 2:1 in the presence of 2 mole % Cp 2TiC12 . The results are listed in
Table 25. The catalyst used in this study was titanocene dichloride,
Cp 2TiC12 , because it worked so well in the previously studied hydro-
metallation reaction with bis-dialkylaminoalanes. All of the hydrides
provided similar results. When allowed to react with terminal alkenes,
the reactions were over in 10 minutes at room temperature. After
quenching with D 20, the products •were analyzed via mass spectrometry.
The mass spectrum of the products were compared to the mass spectrum of
the deuterated species obtained from quenching the corresponding Grignard
Reagents with D 20. The % deuterium incorporation in the product was
considered an indication of the intermediate formation of hydrometallated
species. All terminal alkenes (1-octene i-hexene, styrene, methylene-
cyclohexane, 2-ethyl-l-hexene and neohexene) provided high yields of
deuterium incorporation (95-100%) except for the neohexene reaction which
106
only showed 55% deuterium incorporation. As for the bis-dialkylamino-
alumination reaction discussed above, steric reasons are probably re-
sponsible for this result. Another observation in this study, which
parallel the bis-dialkylaainoalunination reaction is that the overall
yield of the alkane decreased in the following manner: octane (98%) A ,
hexane (99%) > ethylbenzene (85%) > methylcyclohexane (70%) s\, 2-ethyl-
hexane (70%) > neohexane (60%).
The internal alkenes (cis-2-octene, cis-2-hexene amd trans-2-
hexene) once again showed great difficulty in becoming hydrometallated.
In all cases only low yields (1-5%) of alkanes were observed. The
reactions were carried out as described above with the reaction time in-
creases to 36 hours. The reaction temperature was also increased to 60 °C
but only small amounts of alkanes were observed (1-5%). Sato 19 found
that the internal alkenes could be hydroaluminated if the reactions were
carried out for 120 hours at 60°C and in the presence of TiC14 or ZrC14 .
We were hoping that Cp 2TiC12 would provide milder conditions, but un-
fortunately this did not happen.
For terminal alkynes, (1-octyne and phenylethyne) the same problems
associated with the bis-dialkylaminoalumination reactions namely the
deprotonation of the acetylenic proton, and further addition were very
prevalent in these reactions (See Scheme 7). Only small amounts of
alkenes (3-10%) and alkanes (4-13%) were observed. The starting alkyne
showed a great amount (85%) of deuterium incorporation when the reaction
was quenched with D 20. By lowering or raising the temperature, no major
effect was observed except that the rate was affected in the expected way.
Normally, all reactions were conducted at room temperature for 2 hours in
107
a reagent to alkyne ratio of 1:2 in the presence of 2 mole % of Cp 2TiC12 .
By lowering the temperature to 0 °C, after 2 hours, only a trace of alkene
or alkane was observed. After 12 hours, 10% (the same amount observed at
room temperature) alkene and alkane were observed. Longer reaction times'
showed no improvement. By raising the temperature to 50 °C, the amount of
alkane increased to 30% after 30 minutes. After 2 hours 40% octane and
45% octene were observed. If the reaction was quenched with D 20, the
product incorporated only 50% deuterium. It was noticed by earlier
workers in this group that 1-octyne and phenylethyne could be reduced to
1-octene (99%) and styrene (94%) in high yields by LiA1H452
with catalytic
amounts of NiC12 . Unfortunately deuterium analysis was not conducted.
Therefore the author repeated this experiment and quenched it with D 20
after 48 hours at room temperature in THF. The amount of deuterium in-
corporation was only 15%.
The internal alkynes (4-octyne, 2-hexyne, phenylpropyne and 1-tri-
methylsily1-1-octyne) were allowed to react under the conditions described
above. 4-Octyne and 2-hexyne produced 99% and 100% cis-4-octene and cis-
2-hexene respectively with 100% deuterium incorporation for both. The nmr
spectrum of the product from 2-hexyne showed a 53:47 ratio of 2-deuterio-
to 3-deuterio-cis-2-hexene. When this reaction was quenched with a
benzene solution of iodine, 95% of the corresponding vinyl iodo-compounds
were obtained. For the 2-hexyne reaction, a 51:49 ratio of 2-iodo- to 3-
iodo-cis-2-hexene was obtained. For the phenylpropyne case, 15% 1-phenyl-
propane, 70% of cis-I.-phenyl-1-propene and 15% 3-phenyl-1-propene were ob-
tained under these same conditions. The 3-phenyl-1-propene contained 85%
deuterium and the phenylpropane contained only 55% deuterium. These
108
results indicate potential for the production of vinylaluminates which
have been shown to undergo further reaction with CO 2' cyanogen and
halogens to form a,S-unsaturated nitriles and vinyl halides.52
As noted earlier, the terminal octyne reactions were not clean.
Therefore, 1-triraethylsilyl-1-octyne was prepared by the reaction of 1-
octynyllithium with chlorotrimethylsilane as described earlier. When this
alkyne was allowed to react under similar conditions, the reaction was
slow even at 60°C. The best results were a 35% yield of cis-1-trimethyl-
silyl-1-octene (65% deuterium incorporation) and a 15% yield of 1-tri-
methylsilylhexane (20% deuterium incorporation).
In conclusion, these hydroaiumination reactions work very well for
terminal alkene and internal alkynes but not for internal alkenes and
terminal alkynes. An interesting observation which is a result of this
work is that the regiochemistry involved for both the bis-dialkylamino-
alane and aluminate reactions is approximately the same. This indicates
that the regiospecific step must be the same for both reactions. There-
fore it is logical to suggest that the complexation of the intermediate
titanium hydrido compound or formation of the alkyltitanium compound is
the rate determining step after which the subsequent steps of Schemes
1,2 and 3 follow.
Reductions Using HCo(C0) 4
Because of the nature of our research, we thought it would be
appropriate to continue our search for stereoselective reducing agents by
investigating the reduction of ketones, aldehydes, enones and alkyl
halides by transition metal hydrides such as HCo(C0) 4 (easily prepared
from dicobalt octacarbonyl). Dicobalt octacarbonyl [Co(C0) 4 1 2 , was
109
prepared according to a procedure by Orchin.53
Cobalt carbonate was
2 CoC03
+ 2 H2
+ 8 CO [Co (C0)4 2 + - 2 11
20 + 2 CO
2
heated to 160°C (with hydrogen and carbon monoxide in an autoclave
pressurized to 3,500 psi) for 3 hours. Dicobalt octacarbonyl prepared
in this manner was first allowed to react with distilled pyridine to form
the pyridine adduct which after further reaction with concentrated sul-
furic acid produced cobalt hydrocarbonyl, HCo(C0) 4 . The hydride formed
3 [Co(C0) 4 ] 2 + 12 C 5H5N 4. 2 [Co(C 5H5N) 61[Co(CO) ] 2 + 8 CO
2 [Co(C 5H5N) 6][Co(C0) 4] 2 + 8 H2 SO4 4- 4 11C0(C0) 4 + 2 CoSO4 + 6(C5H5NH) 2 SO4
at 0°C and was distilled quickly by bubbling CO through the solution and
trapped in a liquid nitrogen trap; however, carbon monoxide also condensed
under these conditions. The CO was then allowed to evaporate slowly
leaving behind the HCo(C0) 4 which was dissolved in hexane and stored at
-78°C.
Hydrogenations of organic compounds are very well documented under
"oxo process" conditions.54,55
Goetz and Orchin56
have reported that
when a,3-unsaturated aldehydes and ketones are treated with HCo(CO)4 under
"oxo" conditions, the olefinic linkage is reduced rather than hydro-
formylated. They also showed that: HCo(C0) 4 in hexane reacted with
saturated aldehydes at room temperature to give the corresponding alcohols
at a rate much slower than the reactions with olefins under the same
110
conditions. Thus it might be suspected that the reduction of saturated
ketones would also be slow. However, we wanted to observe the stereo-
selectivity of HCo(C0) 4 towards 4-t-butylcyclohexanone and 2-methylcyclo-
hexanone under various conditions (Table 26). We found that the reduction
at -20°C, 0°C, RT and 45 °C for 12 hours produced only 0-5% of the corre-
sponding alcohols with an axial to equatorial ratio of 53:47 and 60:40,
respectively. When the reactions were carried out at 45 °C, decomposition
of HCo(C0)4 became very rapid. At the other temperatures, the reduction
of the ketones took place very slowly. It was anticipated that this re-
duction would be stereoselective; however, this was not the case.
Evidently, the reaction gees through a pathway suggested by Orchin
and Goetz56
for aldehydes and oc,-unsaturated aldehydes and ketones
(Scheme 10). A. semiquantitative kinetic study of these reagents in
which the hydrocarbonyl disappearance was measured by a simple titration
technique discussed in the Analytical Section indicates that the reactions
are first order with respect to hydrocarbonyl. The hydrocarbonyl appears
to function as a hydride donor.
The electron-releasing properties of alkyl groups should help to
stabilze the complex A (Scheme 10) since cobalt is electron withdrawing.
Therefore as the stability of the complex increases the reactivity of the
subsequent reducing step decreases and thereby lowering the yield of
alcohol as was observed.
Simple and Complex Metal Hydride Carbonylations
There is currently intense interest in catalytic reactions of
carbon monoxide and hydrogen. Mechanistic considerations suggest that
formation of the first C-H bond may be the key step. Casey and Neumann57
R slow
HCo(C0) 4 C.:70
/// R '
111
R7
Co(C0) 4
c===o ! co
RI) 7 o --CO
CO
HCo(C0) 4
R2CH(OH) + Co
2(CO)
8
Scheme 10: Proposed Mechanism for the Reduction of Ketones with HCo(CO)4.
112
have suggested that insertion of CO into a metal hydrogen bond is not a
feasible transformation. An alternative would be nucleophilic attack by
a hydride on a coordinated carbonyl. Such reactions lead to formyl, 58
hydroxymethyl 59 and methyl60.61 complexes using boron hydride reagents.
catalytic process involving such a step would require a transition metal
hydride capable of exhibiting hydridic character, which has been observed
only for several complexes of titanium62 and zirconium. 63 Examples of
the reduction of CO to methane64
or to alcohols 65,66 have recently been
reported.
Bercaw, et. al. 65 used metal clusters, 0s 3 (C0) 12 and I 4 (C0) 12 to
catalyze the reduction of carbon monoxide to methane at 140°C and (1, 2
atmospheres. They also observed stoichiometric H 2 reduction of CO to
methoxide under mild conditions with (n5-C5Me
5)2Zr(CO) 2. The inter-
mediate compound was determined to be (n5-c5Me5 ) 2Zr(H)(OCH3) which when
hydrolyzed with aqueous HC1 leads to(n5-05Me5 ) 2ZrC12 , H2 and CH3OH.
Huffman, et al. 64 reported a homogeneous hydrogenation of carbon
monoxide to methane using Cp 2Ti(C0) 2 . When the titanium carbonyl complex
was treated with a mixture of hydrogen and carbon monoxide (3:1 mole
ratio, 1 atm at 25°C) at 150°C, methane was produced. However, the
reaction was not catalytic in titanium.
Schwartz and Shoer66 reported that i-Bu 2A1H (DIBAH) in the
presence of Cp 2ZrCl2 as a catalyst will reduce CO at room temperature
and 1-4 atm to give, on hydrolysis, a mixture of lin.earaliphatic alcohols
(equation 6). The DIBAH rapidly forms a complex A with Cp 2ZrCl2 which
/11----Al\Bu2
Cl
\ II AlBu2
113
A (6)
(7)
(Al) 0 (Al) +
chain propagation
H—(A1)
(Zr)--411 0 (Al)
CO etc.
Cp 2ZrC12 + 3 DIBAH
1) CO 2) H al"
3
CH3 OH + CH
3 CH
2 OH + CH
3CH
2CH
2OH +
can readily absorb CO and upon hydrolysis provide a mixture of alcohols.
They suggested a mechanism in which DIBAH dissociated from A to provide
a vacant coordination site on Zr. After CO was coordinated to the Zr,
reduction of CO could occur by H migration from Zr or by attack by Al—H
to form compound. B. The chain propagation or termination steps were
suggested to occur according to equation 7.
chain termination
(Zr) H + (A1)---fH 0 (Al)
CO
114
Rathke and Feder 67 have also reported the production of alcohols
(methanol, ethanol and propanol) and fotmates (methyl and ethyl) using CO
and hydrogen in the presence of cobalt octacarbonyl in p-dioxane at 182 °C
and 300 atm presumably due to the production of HCo(CO)4 which can react
with CO and incipient radicals.
Based on this background and our work, we thought that it would be
possible to accomplish the above results using CO and simple or complex
metal hydrides (a source of hydrogen, H 2) in the presence of a transition
metal halide catalyst such as Cp 2TiC12 .
When we first started this work, we allowed lithium aluminum
hydride, sodium aluminum hydride, alane, sodium hydride and activated
lithium hydride in the presence of 10 mole percent of various transition
metal halides (Cp 2TiC12 , TiC14 , NiC12 , CoC12 and FeC13 ) in THE or benzene
to react under 4000 psi of 99.5% carbon monoxide obtained from Matheson
Gas Products (Table 27). The only product observed was methyl formate.
However, upon further investigation it was discovered that methyl formate
was present in the carbon monoxide cylinder. It may be safe to assume
that the methyl formate was formed inside the iron cylinder under pres-
sure because the methyl formate was not listed as an impurity by the
company. This reasoning raises the hope that methyl formate can be
prepared from CO in the presence of a catalyst and hydrogen.
According to the literature, numerous reports68
describe the pre-
paration of methyl formate from CO and H2 with metal catalysts and/or
base in small to modest yields. However, we have as yet not been able
to accomplish this transformation. During our investigation we were also
hoping to observe methanol, ethanol, propanol, ethyl formate, formic acid
or acetone which have all iieen reported in the literature as minor pro-
ducts, but these have also escaped our detection (Table 27). Another
possibile product would be methane, but we were not set up to detect it
as a reaction product and therefore are not sure whether or not it was
formed in the above reactions.
115
CHAPTER IV
CONCLUSION
During our investigation, we discovered that terminal alkenes and
internal alkynes are reduced rapidly and in quantitative yields by
HAl(NR2 ) 2 , LiA1H4 , NaA1H4 , LiA1Me3H, NaA1Me3H, LiA1H2 (NR2 ) 2 , NaA1H2 (NR2 ) 2
and Vitride, NaA1H2 (OCH2CH2OCH3 ) 2 in the presence of a catalytic amount
of Cp 2TiC12 . When these reactions were quenched with D 20 or I2 ,excellent
yields of the corresponding deuterium or iodo compounds were obtained in
most cases. However, when benzaldehyde or benzophenone were allowed
to react with bis-dialkylaminoalanes, the corresponding tertiary amines
were obtained in excellent yields. This result may also indicate that
alcohols may be converted to tertiary amines; however, this possibility
was not tested.
Internal alkenes and terminal alkynes did not react rapidly or
regioselectively when mixed with Al—H compounds in the presence of
Cp2TiC12 . However, for the terminal alkynes this situation was remedied
by preparing the corresponding 1-trimethylsilyl derivative which reacted
under hydrometallation conditions to provide the hydrometallated species
in good yield. The trimethylsilyl group could then be removed by acid.
Unfortunately, longer reaction times and higher temperatures were needed.
The reactions of carbonyl compounds, alkenes, alkynes and enones
with activated LiH in the presence of transition metal halides were also
116
117
investigated. However, ieduction of the starting materials was accomp-
lished only when stoichiometric amounts of VC13 were added to the reaction
mixture. Reduction of a representative carbonyl compound, 4-t-butylcyclo-
hexanone, produced 82% of the axial alcohol in 86% yield. Aldehydes were•
reduced to their respective alcohols in high yields (95-97%). Esters
were reduced to the alcohols in high yields (93-95%) with small amounts
(5-7%) of the aldehydes produced as well. Alkynes did not react. The
only a,(3-unsaturated carbonyl compound to react was cinnaldehyde which
produced 90% of the 1,2 reduced product only. Only the terminal olefins
reacted under these conditions. 1-Octene was reduced in 77% yield with
LiH in the presence of Cp2TiC12 .
When simple and complex metal hydrides were allowed to react with
CO under high pressure in the presence of Cp 2TiC12 , TiC1 4 , NiC12 , CoC12
or FeC13, no products other than starting material were detected.
The reduction of ketones by HCo(CO) 4 was also investigated. It
was observed that ketones react very slowly with HCo(C0)4 and with very
little stereoselectivity.
This project encompassed many aspects of hydrometallation reactions
and it has generated more interesting investigations, such as the for-
mation of tertiary amines from aldehydes, ketones or alcohols and the
continuing investigations in the areas of carbonylation of metal hydrides
and the non-isomerized hydrometallated internal olefins.
During our investigations, we discovered an excellant hydrametal-
lation system which consists of bis-dialkylaminoalanes, HAl(NR2 ) 2 , in the
presence of a catalytic amount of transition metal halide [e.g.. bis-
(cyclopentadienyltitanium dichloride, Cp 2TiC12]. Since HA1(NR2 ) 221
cam-
pounds can be prepared by the reaction of aluminum metal, hydrogen and
118
dialkylamine in a one step taction in quantitati4/e yield, and since the
resulting compounds are soluble in hydrocarbon solvents as well as
ethers, these hydrometallating agents should be both versatile and
economically attractive.
119
Table 8. Reactions of 1-Octene with HAl(NPr) 9 the Presence of 5 Mole Percent Catalyst and Quenched with'-D;O.
CATALYST OCTANE
c % DEUTERIUM
(%) b INCORPORATION
TiC13
95 65
TiC14 97 80
VC13
10 0
CrC13
5 0
MnC12 3 0
FeC12 5 0
FeC13
7 0
CoC12 99 10
NiC12 99 10
Cp2TiC12 99 93
Cp2ZrC12 5 95
CpNi(dep)C1 99 5
Ni(acac) 2 5 0
Ni(PEt3 ) 2Br 90 7
Allyl-Ni(dep)Br 81 10
Polymer bound Benzyl- titanocene dichloride 99 0
Cp2VC12 15 2
CuI 5 0
ZnBr2
5 0
a) All reactions were carried out in benzene at RT for 30 minutes under an argon atmosphere.
b) Yield of octane (d 0 + d1 ) was determined by glc using hexane as the internal standard.
c) Percent deuterium incorporation = d1 /(d0 + d1 ) X 100 as determined by mass spectroscopy.
120
Table 9. Reactions of 1-Octene with HA1(NPr2) 2 in the Presence of 5 Mole
Percent Cp2TiC1
2 and Quenched with D
2O.
a
TEMPERATURE (°C)
TIME (min) SOLVENT ATMOSPHERE
b (%) DEUTERIUM- INCORPORATION
RT 60 THE N2 78
RT 60 TIT Ar 87
RT 60 Benzene N2 88
RT 60 Benzene 93
40 10 THE N2 75
40 10 THE Ar 85
40 10 Benzene. N2 88
40 10 Benzene Ar 93
a) The yield of octane was 99% in all cases.
Percent deuterium incorporation = d /(do + d1 ) X 100.
0=
Table 10. Reactions of Alkenes with HAl(NPr 1-2 ) 2 and 5 Mole Percent of Cp 2TiC1 2 and Quench6d with D 20. a
HYDROLYSIS b % DEUTERIUM c- ALKENE PRODUCT (%) INCORPORATION
d 1-Octane Octane (99) 93
cis-2-hexene Hexane (99) 83
trans-2-hexene Hexane (99) 81
121
C.) (") 20
0- (99) 72
(1).- (trace)
t- BuCH C13 (99) 10
Et CH C
\ (90) 75
H 3 I Bu
2-Methylbutane (trace)
Me Me
\. /
H— C C-H (0)
Me Me
t Bu."----\ II
Et 112C "="- C
j 3
CH3 He Me /
Me
H
-- C = CH2 • CH2 •
CH3 (1001' 96
e
Table 10. (continued)
a) All reactions were carried out in benzene at 60 °C for 12 hours.
b) Yield was determined by glc using hexane as the internal standard except for cis- and trans-2-'hexane, neohexene and tetramethylethylene when octane was used. The only other compound observed was that of starting material, if any.
c) Percent deuterium incorporation = d1 /(d0 + d1 ) X 100 as determined by mass spectroscopy.
d) Reaction was over in 15 minutes ar RT.
e) 90% of the deuterated ethylbenzene was determined to be PhCH(D)CH3 with PhCH
2CH2D accounting for the other 10%.
122
123
Table 11. Regioselectivity in the Reaction of B1(NPr2) 2 with Alkenes in
the Presence of 5 Mole Percent Cp 2TiC12 as Determined by
Quenching with a Benzene Solution of Iodine. a
AT KENE PRODUCT (%) b
1-Octene 1-Iodooctane (80)
1-Hexene 1-Iodohe-xane (80)
C is-2-hexene 1-Iodohexane (75)
Trans-2-hexene • 1-Iodohexane (75)
3-Hexene 1-Iodohexane (72) 2 4- 3-Iodohexane (5)
a) All reactions were carried out in benzene for 24 hours at 58 °C except for 1-octene which was complete in 15 minutes at RT.
b) Yield was determined by g1c using dodecane as the internal standard.
THF 40
60 41
Benzene 42
58 45
TEE or
Benzene 45 1‘.."./
55 45
THF 10 80 75
10
Benzene 10 90 90
Table 12. Reactions of Dienes with HAl(NPr i)2 in THF or Benzene at. Room .2 K , Temperature fOr 12 Hours in a Diene/HAI(NPr
2 ) 2 Ratio of 1:2 and Quenched with D2 0.
DIENE' RECOVERED % DEUTERIUM
SOLVENT DIENE (%) a PRODUCTS (%)1) INCORPORATION
124
a) Yields were determined by glc with octane as the internal standard.
b) Yields of products (d + d 1 + d2 + ) was determined by glc using octane as the interna9 standard.
c) By nmr and mass spectroscopy, percent deuterium incorporation = d1/)d
0 + d l ) X 100.
d) Percent deuterium incorporation = d2 /(d0 /(d + d1 + d2) X 100 as deter-
mined by mass spectroscopy.
125
Table 13. Reactions of Cis-2-Heltene and HAl(NPr2) 2 with Vtrious Catalystsa
and Quenched with a Benzene Solution of Iodine.
CIS-2- 'TRANS-2- .HEKENE . (%) —HEKANE . (%) -
1-IODO 2-IODO . BEKANE . (%) - HiKANE(%) CATALYST . REXENE (%)
TiC13
20 15 10 55 0
TiC14
19 13 12 56 0
VC13
93 2 0 0 0
CrC13
93 2 5 0 0
MnC12
95 3 2 0 0
FeC12 70 20 10 0 0
FeC13 75 13 12 0 0
CoC12
0 1 99 0 0
NiC1 2 0 1 99 0 0
Cp2ZrC1 2
90 2 5 3 0
Cp2TiC1
2 8 12 5 75 0
Cp 2TiC1 2d 85 15 trace 0 0
CpNi(Ph3P)C1 3 0 97 trace 0
Ni(acac) 2 90 0 10 0 0
Ni(PEt3
)2 Br2
10 0 90 0 0
Allyl Ni(dep)Br 15 0 85 2 0
Cp2VC1 2 90 3 7 .0 0
a) 5 Mole % in benzene b) All reactions were carried out in benzene at 60 °C for 24 hours. c) Yields were determined by glc with octane as the internal standard and
normalized by % cis-2-hexene + trans- 2-hexene + hexane +.1-iodohexane + 2- iodohexane = 100%.
d) Reaction was carried out in benzene or THE at RT for 24 hours which was monitored every 3 hours by glc.
Table 14. Reactions of Alkynes with M1(N134) 2 and 5 Mole Percent Cp 2TiC12 in an Alkyne/Alane Ratio of 1.0:1.02. a
ALKYNE WORK UP
PRODUCTSb % YIELD
1 -Octynec
D20 Octane-d
0 + Octane-d1 + Octane-d 2 + Octane-d3 54
45 22 18 15
1-Octene-d0 + 1-Octene-d 1 46
22 78
1 -Hexyne c
Hexane-d0 + Hexane-d
1 + Hexane-d
2 + Hexane-d3 49
41 23 19 17
1-Hexene-d0 + 1-Hexene-d1 51
2 -Octyne
18 82
D20 Octane-d0 10
Cis-2-Octene-d0 + Cis-2-Octene-d i 90
3 97
Trans-2-Octene Trace
PRODUCTS % YIELD
5
Cis-2-Rexene-d1 94
96
1
I H 82
Pr
ALKYNE WORK UP
2 -Hexyne Hexane
Cis-2-Hexene-d0
4
Trans-2-Hexene
I9 5C:=C\ Pr Me
Table 14. (Continued)
53 47
PhrCEEt-CH 3 D20 1-Phenylpropane
Ph lie Ph D\' ---Cc: //C. e
10 90
e Ph - ,› 11 EEk:
95
1
Table 14. (Continued)
ALKYNE
WORK UP
PRODUCTS
% YIELD
Hexyl—CH2CH2SiMe 5
Hexyl e Hexyl 3
10 90
87
95
Ph—C---C Ph
Ph p (H)
H Ph
Ph
a) All reactions carried out in benzene at RT for 1 hour and quenched with D20 or a benzene solution
of iodine. b) Yields were determined by glc and are based on alkyne and/oroctane as the internal standard. The
relative ratios of isomers were determined by NKR using benzhydrol as the internal standard. c) Reaction carried out at 0 °C for 8 hours. Reaction carried out at 45 C for 12 hours.
Table 15. Reactions of 2-Hexyne with HAl(NPr 2) 2 and Cp2TiC12 in 1.0:1.02:0.1 Mole Ratio. a
TIME 2-HEXYNE (hr) RECOVERED (%)
HYDROLYSIS PRODUCTS HEXANE (%) CIS-2-HEXENE (%)
TRANS-2- , c
D INCORPOR- ATION (%) HEXENE (%)
0.7 74 7 17
1 72 8 20 95 2 62 8 30 trace 96 4 53 8 39 trace 95
8 30 8 Ag trace 95 16 15 10 73 2 91 16 0 23 47 In 57
a)
b)
c)
Reactions were carried out in benzene at sphere and quenched with D 20.
Yield was determined by glc using octane and normalized (%) 2-hexyne + (%) hexane trans-2-hexene = 100%.
Percent deuterium incorporation = d1/(d
0 by mass spectroscopy.
0°C under an argon atmo-
as the internal standard + (%) cis-2-hexene + (%)
+ dl ) X 100 as determined
d) The temperature was allowed to increase to room temperature.
129
130
Table 16. Reactions of 1-Octene with HAl(NPr )2 -CDt2 TiC1 2: Effect of
2. Temperature and Catalyst Concentration.
MOLE % TEMPERATURE TIME
Cp2TiC12 (°C) (hr) OCTANE (%) b D INCORPORATION (%)c
5 58 2 100 88
10 58 1 100 83
20 58 0.25 100 86
5 25 12 100 86 d
5 25 1.5 10
a) Reactions were carried out in benzene under a nitrogen armosphere and quenched with D20.
b) Yield was determined by glc using hexane as the internal standard.
c) Percent deuterium incorporation = d1/(d
0 + d
1) X 100 as determined
by mass spectroscopy.
d) The percentage of deuterium incorporation increased to 93 when the reaction was conducted under an argon atmosphere and was complete in 15 minutes which was a distinct improvement over the nitrogen experiments.
Table 17. Reactions of the Hydrometallated Species with Carbonyls or Oxygen or Carbon Dioxide in Benzene
at Room Temperature for 24 Hours. a
REAGENT ADDED CARBONYL
RECOVERED CARBONYL (%)
CARBONYL REDUCED ADDITION OTHER
CARBONYL (%) PRODUCT (%) PRODUCTS (%)
• HAl(NPr i) 2 2
Acetone
Benzaldehyde
Benzophenone
(35)
(0)
(0)
'(40) Isopropanol
(100) Benzyl Alcohol
(100) Benzhydrol
(THF) (30) (45) Axial Alcohol (18) Eq. Alcohol (82)
(Benzene) (30) (45) Axial Alcohol (30) Eq. Alcohol (70) --
HAl(NPri) 2 2
Acetone (70) (5) 2
Benzaldehyde (0) (10) Trace (PCH2NP4 (90) 1-Octene
Benzophenone (0) (35) Trace (P 2CHNP4 (65)
HAl(NEt 2)2 Acetone (65) (5) 2
Benzaldehyde (0) (10) Trace (PCH2NEt 2 (90) 1-Octene
Benzophenone (0) (38) Trace (f)2CHNEt
2 (62)
Table 17. Continued)
CARBONYL ADDED RECOVERED REDUCED ADDITION OTHER
REAGENT CARBONYL CARBONYL (%) CARBONYL (%) PRODUCT (%) PRODUCTS (%)
HAl(NPr1) 2 Acetone (66) (5) 1
+ Benzaldehyde (0) (10) 3 49CH2NPr2 (90) 1-Octene
Benzophenone (0) (37) 3 th2 CHNPr 2i (63) '
HAl(NP4) 2 02
No Reaction
CO, No Reaction
Yields for the acetone reactions were determined by glc based on added carbonyl. Yields for the benzaldehyde and benzophenone reactions were determined by NMR using acetone as the internal standard.
133
Table 18. Reactions of 1-Octene with Substituted Alanes in the Presence of 5 Mole % Cp2TiC12 in Benzene at Room Temperature for 12 Hours.
Alane Work Up Product (%) a
HAl(NPr2)2 I2 1-Iodooctane (80)
D20 Octane-d 1 (93)
HAl(NEt 2 ) 2
HAl[N(SiMe3 ) 2 ] 2
H2A1C1
1-Iodooctane (86)
Octane-d 1 (90)
I2 1-Iodooctane (53) 2-Iodooctane (47)
:D 20 Octane-d 1 (93)
I2
1-Iodooctane (70)
D20 Octane-d1 (85)
HA1C12 12 1-Iodooctane (68)
D20 Octane-d 1 (83)
H2 A10Meb , c
12 1-Iodooctane (10)
D 20 Octane-d 1 (10)
HA1(0Me)2, c
I2 1-Iodooctane (0)
D20 Octane-d 1
(0Pr 1 ) c2 ' d D20 Octane-d 1 (15)
Hk1(0But ) 2 c,d
D20 Octane-d 1 (15)
a) Yields were determined by glc based upon 1-octene. b) Insoluble c) The reagents were prepared in THF but it was removed by vacuum and
replaced by freshly distilled benzene. This procedure was repeated three times. According to glc, only 5-10% of THF remained.
d) Slightly soluble
134
Table 19. Reactions of LiH and Transition Metal Halide with 4-t-Butyl-cyclohexanone in a 1:1:1 Ratio at Room Temperature for 24 Hours in THF.
TRANSITION AXIALa
EQUATORIALa METAL HALIDE ALCOHOL (%) ALCOHOL (%)
YIELD (%)
None 45 55 5
CrC13 83 17 8
MnC12 0 0 0
FeC12 0 0 0
CoC12 0 0 0
NiC12 0 0 0
TiC13 61 39 41
VC13: 82 18 86
FeC13 74 26 68
Cp 2TiC12 65 35 27
VC13b
80 20 3
a) Yields were determined by glc based on internal standard.
b) Only 5 Mole % of VC13 was added. c) When this reaction was allowed to take place in benzene, an
80% yield of the alcohols was obtained with the axial alcohol consisting 79% of the total.
SUBSTRATE
Benzaldehyde
Hexanal
Ethyl Benzoate
Ethyl Butyorate
135
Table 20. Reactions of Carbonyl Substrates with LiH:VC1 3 in THE at 45 °C for 36 Hours in a Mole Ratio of 1:3.
RECOVERED SUBSTRATE (%)
AXIAL OR EXO ALCOHOL (%) a
EQUATORIAL OR ENDO ALCOHOL (%1 YIELD (%),
11 82 18 86
3 90 10 90
92 8 90
3 95 5 95
0 Benzyl Alcohol 97 97
0 Hexanol 95 95
0 Benzaldehyde 7 Benzyl Alcohol 93 95
0 Butanal 5 Butanol 95 95
a) Yields were determined by glc based on internal standard.
136
Table 21. Reactions of Alkenes with LiH:VC13 in THE at 45 °C for 36 Hours
in a Mole Ratio of 1:3 and Quenched with D 20.
SUBSTRATE RECOVERED SUBSTRATE (%) ALKANE (%) a D INCORPORATION (%) b
1-Octene 0 Octane (95) 30
Cis-2-Hexene 100 Hexane (0)
Trans-2-Hexene 100 Hexane (0)
2-Ethyl-1-Hexene 5 3-Methylheptane (95) 29
Cyclohexene 100 Cyclohexane (0)
1-Methyl-l- cyclohexene 100 Methylcyclohexane (0)
Methylene-cyclohexane 7 Methylcyclohexane (93) 30
a) Yields were determined by glc based on internal standards.
b) Percent deuterium incorporation = d 1 /(d0 + d 1 ) X 100 as determined by mass spectroscopy.
Table 22. Reactions of tin and NaH with 1-Octene in the Presence of Catalytic Amounts of Transition Metal Halides in Benzene at Room Temperature for 24 Hours and Quenched with D20.
METAL HYDRIDE
5 MOLE % CATALYST
RECOVERED 1-OCTENE (%) a OCTANE (%) a D INCORPORATION (%)lb
LiH VC13 94 Trace
Cp2TiC12 18 77 50
TiC14 33 59 45
FeCl3 74 25 <5
NiC12 70 27 <5
CoC12 65 32 <5
NaH VC13
97 0
Cp2TiC1 2 94 5
TiC14 97 Trace
FeC13 97 0
NiC12 97 Trace
CoC12 97 Trace _ - - -
a) Yields were determined by gic and based on hexane as the internal standard.
b) Percent deuterium incorporation = d i /(do + dl ) X 100 as determined by mass spectroscopy.
137
138
Table 23. Reactions Of Alkynes with LiH:VC1 1 in Benzene at 45°C for 36 Hours in a Mole Ratio of 1:3 and Iuenched with D 20.
SUBSTRATE RECOVERED SUBSTRATE (%)
ALKANE OR ALKENE (%) D INCORPORATION (%)
b
1-Octyne 100 0
1 -Hexyne 100 0
2 -Hexyne 100 0
1 -Phenylpropyne 100 0
Diphenylethyne 100
Phenylethyne 100 0 0
a) Yields were determined by glc and based on an internal standard.
b) Percent deuterium incorporation = d igdo + d1 ) X 100 as determined by mass spectroscopy.
10 90 0 0
10
100 0 0
100 0 0
Table 24. Reactions of Enones with LiH:VC1 3 in . Benzene at 45 °C for 36 Hours in a Mole Ratio of 1:3.
RECOVERED 1,2-REDUCTION 1,4-REDUCTION TOTAL 'ENONE ENONE (%) TRODUCT'M .a PRODUCT :(7.) a- REDUCTION . M a
139
Ph
Ph
a) Yields were determined by glc based on internal standards.
LiAIH4 1-Octene d Octane (98)
1-Hexend Hexane (99)
Styrene e Ethylbenzene
0=
Table 25. Reactions of Complex Aluminum Hydrides with Olefins and Alkynes in the Presence of 5 Mole % Cp 2TiC12 in THE for Two Hours in a Mole Ratio of I:1 and Quenched with D20.
a
COMPLEK
ALUMINUM UNSATURATED
HYDRIDE HYDROCARBON PRODUCTS (%)b
D'INCORPORATION'af
140
99
100
(70) 100
(70) 95
(70) 95
(60) 55
Cis-2-Hexene Hexane (5)
Hexane (5)
0
Trans-2-Hexene
(3)
(0)
Table 25. (Continued)
COMPLEX
ALUMINUM UNSATURATED
HYDRIDE HYDROCARBON PRODUCTS"(%) " D'INCORPORATION"(%) c.
LiA1H4
I -Octyne Octane (4)
I -Octene (3)
PhenyIethyne
Styrene (10)
Ethylbenzene (13)
4 -Octyne Cis -4 -Octene (99) 100
2 -Hexyne Cis -2 -Hexene (99) 100
Ph-CESCMe PhCH2CH2CH3 ( 1-5) 55
1/11
C—C 4 \„
(70) 95
Ph
me
e
90 10
PhCH2 CH=CH2 (15)
85
Hexyl-e5=91-SiMe3 Hexyl-CH2CH S" e.3 (15) 20
Hex1 Sitie3
77-C\ (35) 65
H H
141
Bu
Table 25. (Continued)
COMPLEX
ALUMINUM. UNSATURATED
HYDRIDE HYDROCARBON PRODUCTS (%)b
D 'INCORPORATION (%)
LiAlMe3
H 1-Octened Octane (98)
100
1-Hexene
Hexane (98)
100
Styrene
Ethylbenzene (73)
94
142
(69) 94
(71) 95
(58) 53
Cis;-2-Rexene
Hexane
(6)
Trans-2-Hexene Hexane (5)
(0)
(0)
143
Table 25. (Continued)
COMPLEX
ALUMINUM UNSATURATED
HYDRIDE HYDROCARBON PRODUCTS (%') b (%)c
LiAlMe3H 1-Octyne Octane (2)
I-Octene (5)
PhenyIethyne Styrene (9)
Ethylbenzene (11)
- -
4-Octyne Cis-4-Octene (100) 100
2-Hexyne Cis-2-Hexene (99) 100
Ph--=—C-Me PhCH2CH2CH3 (17) 51
/11 —C\ ^
Ph ,pe (70) 97
Me
90
10
PhCH2 CH=CH2 (13) 83
• Hexy1-466.=0-SiMe3 Hexyl-CH2CH2SiMe3 (13) 19
Hex
c(: (34) 65
144
Table 25. (Continued)
COMPLEX
ALUMINUM UNSATURATED
HYDRIDE HYDROCARBON PRODUCTS (%) b D INCORPORATION (%)
NaA1Me3H 1-Octened Octane (100) 100
17Hexene Hexane (99) 100
Styrenee Ethylbenzene (71) 98
c)-- (69) 93
Bu
(58) 53
Cis-2-Rexene Hexane (3)
Trans-2-Hexene Hexane (3)
( 0 )
( o ) - _
Table 25. (Continued)
COMPLEX
ALUMINUM UNSATURATED
HYDRIDE HYDROCARBON PRODUCTS (%) D'INCORPORATION . (%)
NaA1lie3H 1 -Octyne Octane (6)
PhenyLethyne
1-Octane (6)
Styrene (9) ••••••
Ethylbenzene (10)
4-Octyne Cis-4-Octene (99) 100
2 -Hexyne Cis-2-Hexene (100) 99
Ph-C=C-Me PhCH2CH2CH3 (13) 51
/I1 1\ ph/c:=C\ +
MeP
(68)
e
96
90 10
PhCH2CH=ZCH2 (19) 86
HexyI-1EED-SiMe3 Hexy1-CH2CH2S1lle3 (17) 20
(30) 68
145
Cis -2 -Hexene
Trans-2-Rexene Hexane
Hexane
51
146
Table 25. (Continued)
COMPLEX ALUMINUM UNSATURATED HYDRIDE HYDROCARBON PRODUCTS (%) b
D INCORPORATION (%)
Vitride 1-Octene d Octane (100) 100
1-Rexene Hexane (100) 100
Styrene e Ethylbenzene (72) 99
(70) 94
i (69) 93 Bu 'Bu
0 (0)
(0)
Table 25, (Continued)
COMPLEX
ALUMINUM UNSATURATED
HYDRIDE HYDROCARBON PRODUCTS (%) D INCORPORATION (%)
Vitride I-Octyne Octane (5)
I-Octene (6)
Phenylethyne Styrene (II)
Ethylbenzene (10)
4-0c tyne Cis-4-Octene (99) 100
2-Hexyne Cis-2-Hexene (100) 100
Ph-CaECHMe PhCH2 CH2CH3 (14) 50
D\ ÷ (69) 93
Ph \As Pfi Me
90 10
PhCH2CH=CH2 (17) 82
Hexyl-C:=G-SiKe3 Hexyl-CH2CH2SiHe3 (18) 16
liMe3 C\
(35) 61
147
Cis-2-HeXene Hexane
Trans-2-Hexene Hexane
148
Table 25. (Continued)
COMPLEX ALUMINUM UNSATURATED HYDRIDE HYDROCARBON - PRODUCTS (%) b
D'INCORPORATION (%)c
NaAIH4 1-Octened
Octane (99) 100
17Hexene Hexane (99) 100
Styrene e
Ethylbenzene (75) 97
(71) 94
Bu
(62) 56
( 0 )
149
Table 25. (Continued)
COMPLEX ALUMINMA UNSATURATED HYDRIDE HYDROCARBON PRODUCTS (70 D'INCORPORATION'(%)
NaA1H4 1 -Octyne Octane (4)
I-Octene (3)
Phenyiethyne Styrene (11)
Ethylbenzene (15)
— —
4-Octyne Cis-4-Octene (99) 100
2-Hexyne Cis-2-Hexene (98) 100
-Me PhCH CH CH 2 2 3 (8) 50
H\ (75)
94 Ph Me
90 10
PhCH2CH=H2 (17)
82
Hexyl-CEED-SiKe3 Hexy1-CH2CH2Sille3 (18) 18
"Nomemm. (39) 67
NaA1H2 (NEt )
00
96
■
93
1 -octene 1 -octyne
4 -octyne
cis -2 -octene
NaA1H2(NiPr
2 1-octene 1-octyne
4-octyne
cis-2-octane
octane (15) 1-octene (15) octane (13) 4-octene (93)
octane (trace)
octane (95) 1-ocetene (10)
octane (13)
4-octene (91)
octane (2)
150
Table 25. (Continued)
COMPLEX ALUMINUM UNSATURATED
HYDRIDE HYDROCARBON PRODUCTS . (7) D INCORPORATION (% )
LiA1H2 (NEt2
)2 1-oct ene octane (96)
98 I-octyne 1-octene (10)
octane (11)
4-octyne 4-octene (95)
97 cis-2-octene octane (5)
LiA1H2 (NiPr2 ) 2 1-octene octane (92)
96 I-octyne 1-octene (12)
octane (11) 4-octyne 4-octene (97)
97 cis-2-octene octane (2)
a) Reaction were carried out in THE at Room Temperature. b) Yields were determined by glc based on internal standards. c) Percent deuterium in-corporation = d
1 /(d
0 + d
1) X 100 as determined by mass spectroscopy. d)
When the reaction was quenched with a benzene solution of iodine, a 95% yield of only 1-iodooctane was obtained. e) 90% of deuterium was deter-mined to be located on the carbon adjacent to the phenyl ring, PhCH(D)CB 3 . f) When the reaction was quenched with a benzene solution of iodine a 95% yield of a 51:49 ratio of 2-iodo to 3-iodo-cis-2-hexene was obtained.
151
Table 26. Reaction of Ketones Kith E.Co(CO) at Various Temperatures in '4
Hexane and Ketone:11Co(C0), of 1:2 4
Ketones Temperature' Axial Alcohol .(%)
Equatorial Alcohol (%) 4 Yielda
-22 53 47 3
0 55 45 2
RT 53 47 3
45 53 47 1
-22 60 40 3
0 59 41 3
RT 59. 41 2
45 59 41 1
a) Yields were determined by glc and based on internal standards.
Table 27. Carbonylations of Simple and Complex Metal Hydrides in the Presence of 5 Mole Percent Transition Metal Halides at 4000 psi, Room Temperature, in THE or Hexane for 20 Hours.
a TRANSITION METAL !IETAL HYDRIDE
MLIDP PRODUCTS
LiH(activated) none
TiC1 4 Ai H 3 C 112TiC1 2 LiA1H4 NCI
' 2
NaA1H4 CoC1 2
FeC1 3
No Reaction
a) Each metal hydride was allowed to react with each transition metal halide studied.
152
LITERATURE CITED
1. R. Noyoi, I. Umeda and T. Ishigami, J. Org. Chem., 37, 1542(1972).
2. G. P. Boldrini, M. Panunzio and A. Umani-Ronchi, Chem. Commun., 359 (1974).
3. T. Mitsudo, Y..Watanabe, M. Yamashita and Y. Takegami, Chem. Commun., 1385(1974).
4. Y. Watanabe, T. Mitsudo, M. Yamashita, S. C. Shiiu and Y. Takegami, Chem. Lett., 1879(1974).
5. Y. Watanabe, M. Yamashita and T. Mitsudo, Tetrahedron Lett., 1880 (1974).
6. G. P. Boldrini, M. Panunzio and A. Umani-Ronchi, Synthesis, 733(1974).
7. H. Alper, Tetrahedron Lett., 2257(1975).
8. R. J. Bolkman, Jr. and R. Michalak, J. Au. Chem. Soc., 96, 1623(1974).
9. S. Masamune, G. S. Bates and P. E. Geroghiou, J. Am. Chem. Soc., 96 3686(1974).
10. G. M. Whitesides, J. San Filippo, Jr., E. Casey, J. Am. Chem. Soc., 91, 6542(1969).
11. T. Yoshida and E-I. Negishi, Chem. Commun
12. H. Alper, J. Org. Chem., 37, 3972(1972).
13. P. C. Wailes and H. Weigold, J. Organometal. Chen., 24, 405(1970).
14. P. C. Wailes, H. Weigold and A. P. Bell, J. Organometal. Chem., 131., C32(1972).
15. D. W. Hart and J. Schwartz, J. Am. Chen Soc., 96, 8115(1974).
16. P. C. Wailes, H. Weigold and A. P. Bell, J. Organometal. Chem., 373(1971).
17. D. W. Hart, T. F. Blackburn and J. Schwartz, J. Am. Chem. Soc., 679(1975).
18. E. L. Muetterties, Ed.,"Transition Metal Hydrides", Marcel Dekker, Inc., New York, NY, 1971.
153
R. Stedronsky and C. P.
., 762(1974).
154
19. a) F. Sato, S. Sato and M. Sato; J. Organotetal.'Chem.; 131 C26(1977). b) F. Sato, S. Sato and M. Sato; J. Orgatotetal . Chen.;122, C25(1976). c) F. Sato, S. Sato, H. Kodama and M. Sato; J. OtganOMetaU .Chem., 142, 71(1977).
20. E. C. Ashby, J. J. Lin and A. B. Goel, J. Org. Chem., 43, 1263(1978). •
21. R. A. Kovar and E. C. Ashby, Inorg. Chem., 10, 893(1971).
22. F. W. Walker and E. C. Ashby, J. Chem. Ed., 45, 654(1968).
23. D. F. Shriver, "The Manipulations of Air-Sensitive Compounds", McGraw-Hill, New York, NY, 1969.
24. H. Gilman and A. H. Harbein, J.. Am. Chem. Soc., 66, 1515(1944).
25. R. A. Benkeser and R. A. Bickner, J. Am. Chem. Soc., 80 5298(1958).
26. E. C. Ashby, J. R. Sanders, P. Claudy and R. Schwartz, J. An. Chem. Soc., 95, 6485(1973).
27. D. L. Schmidt and E. E. Flagg, Inorg. Chem., 6, 1262(1967).
28. a) E. C. Ashby and R. A. Kovar, Inorg. Chem., 10, 893(1971). b) E. C. Ashby, P. Claudy and R. D. Schwartz, Inorg. Chem., 13, 192(1974).
29. a) T. S. Piper and G. Wilkinson, J. Inorg. Nucl. Chem., 3, 104(1956). b) H. Chatt and B. L. Shaw, J. Chem. Soc. A, 1718(1960). c) R. J. H. Clark, Comprehensive Inorganic Chemistry, 3, 392ff. ( 1 973).
30. I. Wender, H. W. Sternberg, S. Metlin and M. Orchin, Inorganic Synthesis, Vol.5, 190(1947).
31. H. W. Sternber, F. Wender and M. Orchin, Inorganic Synthesis, Vol.5, 192(1947).
32. H. W. Sternber, I. Wender and M. Orchin, Anal. Chem., 24, 174(1952).
33. F. A. Carey and J. R. Toler, J. Org, Chem., 41 1966(1976).
34. G. Weilp and C. C. Whitney, J. Am. Chem. Soc., 89, 2753(1967).
35. J. E. Bercaw and H. H. Brintzinger, J. Am. Chem. Soc., 91, 7301(1969).
36. E. C. Ashby and S. A. Noding, Tetrahedron Lett., 4579(1977).
37. Discussion with Dr. Barefield.
38. J. W. Lauker and R. Hoffmann, J. Am. Chem. Soc., 98, 1729(1976).
39. D. W. Hart, Ph. D. Thesis, Princeton University, 1975.
40. L. I. Zakharkin and O. I. Okhlobystin, Bull. Acad. Sci. USSR, 1236 (1958).
41. a) A. Reinacker, Angew. Chem. Int. Ed., 6, 872(1967). b) K. Ziegler, Angew. Chem., 68, 721(1956).
42. J. J. Eisch and W. C. Kaska, J. Organometal. Chem., 2, 184(1964).
43. E. C. Ashby and J. R. Boone, J. Org. Chem., 41, 2890(1976).
44. E. C. Ashby and J. Laemmle, Chen. Rev., 75, 521(1975).
45. O. Mitsunobu, M. Wada and T. Sano, J. Am. Chem. Soc., 94 679(1974).
46. K. Ziegler, R. Koester and W. R. Kroll, Ger. 1,165,003(C1.0 07c).
47. A. Allred and H. Rochow, J. Illorg. Nuclear Chem., 5, 264(1958).
48. G. Guillaumet, L. 43(1975).
49. G. Guillaumet, L. 353(1977).
50. J. J. Brunet, L. Lett., 1069(1977)
Mordenti and P. Caubere, J. Organanetal. Chem., 92,
Mordenti and P. Caubere, J. Organometal. Chem., 102,
Lordenti, B. Loubinoux and P. Caubere, Tetrahedron 51. J. J. Brunet and P. Caubere, Tetrahedron Lett., 3947(1977).
52. a) G. Zweifel and C. C. Whitney, J. Am. Chem. Soc., 89, 2753(1967). b) G. Zweifel and R. B. Steele, J. Am. Chem. Soc., 89, 2754(1967). c) G. Zweifel, J. T. Snow and. C. C. Whitney, J. Am. Chem. Soc., 90, 7139(1968).
53. M. Orchin, "Advances in Catalysis, Vol. V", Academic Press, New York, NY p. 387(1953).
54. L. Marko, Chem.'Ind., 260(1962).
55. L. Marko, Troc.'Chem. Soc., 67(1962).
56. R. W. Goetz and M. Orchin, J. Org. Chem., 27, 3698(1962).
57. C. P. Casey and S. M. Neumann, J. Am. Chem. Soc., 98, 3595(1976).
58. A. N. Nesmeyanov, K. N. Anisimov, N. E. Kolobova and L. L. Krasnoslo-bodskaya, Izv. Akad. Nauk. USSR, Ser. Khim., 860(1970.
59. R. P. Stewart, N. O. Kamoto and W. A. G. Graham, J. Organametal. Chem., 42, C32(1972).
155
156
60. P. M. Treichel and R. L. Shubkin, Inorg. Chen., 6, 1329(1967).
61. L. J. Guggenberger and F. N. Tebbe, J. An. Chem. Soc., 98, 4137(1976).
62. P. C. Wailes and H. Weigold, J. Organoraetal. Chem., 29, 413(1970).
63. J. Schwartz and J. A. Labinger, Angew. Chem., Int. Ed. Engl., 15, 333 (1976).
64. J. C. Huffman, J. G. Stone, W. C. Krusell and K. G. Caulton, J. Am. Chem. Soc., 99, 5829(1977).
65. J. M. Manriquez, D. R. McAlister, R. D. Sanner and J. E. Bercaw J. Am. Chen. Soc., 98, 6733(1976).
66. L. I. Shoer and J. Schwartz, J. Am. Chem. Soc., 99, 5831(1977).
67. J. W. Rathke and H. M. Feder, J. An. Chem. Soc., 100, 3623(1978) and ref. therein.
68. M. Enderli, U. S. 1,555,796, Sept. 29, 1925; W. Kotowski, N. Paterok, Chem. Anal., 13, 95(1968); D. K. Nandi and B. N. Avasthi, Indian . J. Technol., 5, 266(1967); W. Kotowski, Chem. Technol., 19, 418(1967); L. Kaplan, Ger Offen. 2,559,057 (CI.007C31/18), 08 July 1976, U. S. Appl. 537,885,02 Jan 1975.
PART III
REACTIONS OF MAGNESIUM HYDRIDE:
STEREOSELECTIVE REDUCTION OF CYCLIC AND
BICYCLIC KETONES BY LITHIUM ALKOXYMAGNESIUM HYDRIDES
1.57
158
CHAPTER I
INTRODUCTION
:Background
The use of metal hydrides as stereoselective reducing agents in
organic chemistry has received considerable attention recently.2,3 Al-
though numerous reports have appeared in the literature concerning the
reduction of cyclohexanones by hydrides of boron and aluminum, little is
known about reductions with magnesium hydride and its derivatives pre-
sumably because of the reported lack of reactivity of magnesium hydride
and its insolubility in all solvents studied and also because derivatives
of magnesium hydride were not known. 4 Recently, Ashby and co-workers'
have prepared some THE soluble magnesium-hydrogen compounds of the types
HMgOR5 and HMgNR26 which have been shown to exhibit considerable stereo-
selectivity toward cyclic and bicyclic ketones.7
If HMgOR compounds are
such good stereoselective reducing agents by virtue of their bulky alkoxy
groups, then it would be reasonable to assume that similar "ate" complexes
(e.&. alkali metal alkoxymagnesium hydrides) might produce an even greater
effect.
Purpose
Reactions of tetrahydrofuran soluble lithium alkoxymagnesium
hydrides, LiMgH2 (OR) 8 (where R = methyl, isopropyi4 t -butyl, neopentyl,
cyclohexyl, 2-methylcyclohexyl, phenyl, 2,6-diisopropylphenyl, 2,6-di-t-
buty1-4-methylphenyl, diphenylmethyl, 2,2,6,6-tetramethylcyclohexyl and
159
2,2,6,6-tetrabenzylcyclohexyl) with cyclic and bicyclic ketones such as •
2-methylcyclohexanone, 4-t-butylcyclohexanone, 3,3,57-trinethylcyclohexa-
none and camphor have been 'studied in order to observe any unusual stereo-
selectivity.
160
CHAPTER II
ECPERIMENTAL
General Considerations
Reactions were performed under nitrogen or argon at the bench
using Schlenk tube techniques 9 or in a glove box equipped with a recircu-
lating system using manganese oxide columns to remove oxygen and dry ice-
acetone traps to remove solvent vapors.10
Calibrated syringes equipped
with stainless needles were used for transfer of reagents. Glassware and
syringes were flamed and cooled under a flow of nitrogen or argon. Ketone,
alcohol and internal standard solutions were prepared by weighing the
compound in a tared volumetric flask and diluting with the appropriate
solvent.
All melting points are corrected. The proton NMR spectra were
determined at 60 MHz using a Varian, Model T-60 NMR spectrometer. The
chemical shifts are expressed in ppm (S values) relative to Me4Si as the
internal standard. The mass spectra were obtained with a Hitachi (Perkin-
Elmer) Model RMU-7 or a Varian, Model M-66, mass spectrometer. GLPC
analyses were carried out on an F & M Model 700 or Model 720 gas chroma-
tograph. The it spectra were obtained using a Perkin-Elmer, Model 621 or
Model 257 infrared spectrometer. High pressure work was conducted in an
autoclave (rated to 15,000 psi) obtained from the Superpressure Division
of the American Instrument Company.
"Analyses
Gas analyses were carried out by hydrolyzing samples with 0.1 M
hydrochloric acid on a Standard vacuum line equipped with a Toepler pump. 9
Magnesium was determined by titrating hydrolyzed samples with standard
EDTA solution at pH 10 using Eriochrome-Black T as an indicator. Lithium
reagents were analyzed by the standard Gilman double titration method
(titration of total base followed by titration of total base after reaction
with benzyl chloride) 11 or by flame photometry. The amount of active C-Mg
and C-Li was determined by titrating the active reagent in a dry box with
dry 2-butanol in xylene using 2,2'-diquinoline as an indicator. Aluminum
was determined by adding excess standard EDTA solution to hydrolyzed
samples and then back titrating with standard zinc acetate solution at pH
4 using dithizone as an indicator. Carbon, hydrogen analyses were carried
out by Atlantic Microlab, Inc., Atlanta, Georgia.
Materials
Solvents
Fisher reagent grade anhydrous diethyl ether was stored over sodium
metal, then distilled under nitrogen from LiA1H4 and/or sodium benzo-
phenone ketyl just prior to use.
Fisher reagent grade tetrahydrofuran (liff') was dried over NaA1H 4
and distilled using diphenylmethane as an indicator, under nitrogen, just
prior to use.
Fisher reagent grade benzene and hexane were stirred over concen-
trated H2 SO4' washed with Na2CO3, then distilled water, dried over an-
hydrous MgSO4 and distilled from NaA1H 4 , under argon, just prior to use.
161
162
Preparation of 2,2,6,6-tetramethylcyclohexanone
Into a three-necked, 2-1 round bottom flask, was placed 24.0 g
(0.5 mole) of sodium hydride (50% mineral oil dispersion). The sodium
hydride was washed three times with dry hexane by swirling, allowing the
hydride to settle, and then decanting the liquid portion in order to re
move the mineral oil. The flask was immediately fitted with a magnetic
stirrer, a reflux condenser and a pressure-equilizing dropping funnel
fitted with a rubber serum cap. A three-way stopcock, connected to the
top of the reflux condenser, was connected to a water aspirator and a
source of dry argon. The system was evacuated until the last traces of
hexane were removed from the sodium hydride and then flushed with argon
by evacuating and refilling with nitrogen several times. The aspirator
hose was removed and this arm of the stopcock connected to a mineral oil
filled bubbler. To the dry sodium hydride was added 250 ml of freshly
distilled benzene. Carefully 55 ml (0.5 mole, 44.1 g) of t-amyl alcohol
was added to the stirred mixture. The very vigorous reaction was con-
trolled by the drop rate of the alcohol and by a 5-100C ice-water bath.
After addition of the alcohol, an additional 200 ml of alcohol and 100
ml of benzene were added and the ensuing pale brown solution was warmed to
room temperature. To this solution was first added 32 ml (0.5 mole,
71.0 g) of iodomethane (Aldrich) and then 10.5 ml (0.10 mole, 9.85 g) of
distilled cyclohexanone (Aldrich). The solution turned a bright yellow
color and was allowed to stir over night. The reaction mixture was care-
fully poured into 500 ml of cold water. The mixture was extracted three
times with hexane, washed once with water, dried over anhydrous sodium
sulfate, filtered and the solvent removed by rotary evaporation. The
163
crude product was distilled to give 10.5 g (68% yield), b.p. 50-53°C
(3 mm) [lit. 12 b.p. 62-65 °C (5 mm.)1; IR(neat, film) 2950(s), 2920(m),
2860(m), 1700(s), 1470(s), 13900n), I370(m), I040(m); NMR (CDC13 , TMS)
12H singlet at 1.10 ppm, 6H multiplet at 1.55-1.86 ppm; mass spectrum.,
m/e (rel. intensity) 154(e, 74), 140(7), 11(7), 83(53), 82(76), 78(18),
72(32), 70(13), 69(65), 57(23), 56(100), 55(62), 41(86).
Anal. Calcd for C10H180: C, 77.87; H, 11.76. Found: C, 77.81;
H, 11.70.
Preparation of 2,2,6,6-tetrabenzylcyclohexanone
The same procedure for preparing 2,2,6,6 -tetramethylcyclohexanone
was followed. After the formation of sodium t-amylate, 59.5 ml (0.5 mole,
85.50 g) of benzyl bromide was added followed immediately by 10.5 ml
(0.10 mole, 9.85 g) of freshly distilled cyclohexanone. The solution
turned a grayish white color and was allowed to stir overnight. The
reaction mixture was worked up as described above yielding white crystals
which upon recrystallization from diethyl ether yielded 15.1 g (33%) of
2,2,6,6-tetrabenzylcyclohexanone, m.p. 155-156 °C [lit. 13 m.p. 155-156 °C];
IR (CDC13 , cavity cell) 3060(w), 3040(w), 3000(m), 2930( ), 2860(w),
1685(s), 1600(m), 1490(s), 1450(s), 1250(s), 860(s); NMR (CDC1 3, TMS)
8H singlet at 1.60 ppm, 6H multiplet at 2.22-2.97 ppm, 20H multiplet at
6.90-7.34 ppm; mass spectrum, m/e (rel. intensity) metastable ion 366.5
(e-92, 35), 277(23), 275(10), 224(34), 91(100).
Anal. Calcd for C34H340: C, 89.04; H, 7.47. Found: C, 88.90;
H, 7.45.
164
Purification of Alcohols
Methanol (Fisher) was distilled after treating with magnesium
metal. Isopropanol (Fisher) was distilled over Al(OPr i ) 3 and t-butyl
alcohol (Fisher) was fractionally crystallized under nitrogen. Cyclo-
hexanol, 2-methylcyclohexanol, phenol and 2,6-diisopropylphenol (Ethyl
Corporation) were distilled prior to use. Benzhydrol (Aldrich), neopentyl
alcohol (Aldrich) and 2,6-di-t-butylcresol (Eastman) were used without
further purification.
Preparation of 2,2,6,6-tetramethylcvclohexanol
Into a dry 200 m1 one-necked round bottom flask equipped with a
magnetic stirring bar and a rubber serum cap was placed 10.0 g (0.065 mole)
of 2,2,6,6-tetrauethylcyclohexanone in 30 ml of distilled diethyl ether.
To this stirred solution, cooled by an ice-water bath, was carefully
added 45 ml of 1.50 M solution of LiA1H 4 in diethyl ether. The reaction
mixture was allowed to warm to room temperature and stirred for 2 hours
after which time 20 ml of saturated NH4C1 solution was added carefully
using an ice-water bath to moderate the reaction. The mixture was ex-
tracted three times with diethyl ether, washed once with water, dried
over anhydrous sodium sulfate, decanted and the solvent removed by rotary
evaporation. The crude product was distilled to give 10.0 g (98% yield
of 2,2,6,6-tetramethylcyclohexanol, b.p. 54-55 °C (3.6 mm) [lit. " b.p.
62-65°C (5 mm)1; NMR (CDC13, TMS) 6H singlet at 0.92 ppm, 6H singlet at
0.96 ppm, 6H multiplet at 1.12-1.76 ppm, 1H singlet at 1.92 ppm, IH
singlet at 3.00 ppm; mass spectrum, m/e (rel. intensity) 156(M +, 7),
138(13), 123(18), 109(83), 95(15), 82(100), 69(89), 55(44), 43(56),
41(88); IR (CDC13 , cavity cell) 3500(broad-m), 2945(s), 2920(m), 2855(m),
1460(s), 1385(m), 1370(0, 1030(m), 850(m).
Anal. Calcd for C10H200: C, 76.86; H, 12.90. Found: C, 76.78;
H, 12.83.
Preparation of 2,2,6,6-tetrabenzyLcyclohexanol
The same precedure for preparing 2,2,6,6-tetranethylcyclohexanol
was followed. To 15.0 g (0.033 mole) of 2,2,6,6-tetrabenzylcyclohexanone
in 30 ml of freshly distilled diethyl ether was added 25 ml of 1.5 /4
LiA1H4 in diethyl ether. The product was worked up as described above.
The yield of 2,2,6,6-tetrabenzylcyclohexanol after being recrystallized
twice from diethyl ether, (m.p. I59-160 °C; lit. 15 m.p. 161°C) was 10.0 g
(66%): IR (CDC1 3 , cavity cell) 3560(broad-m), 3060(m), 3040(m), 3010(s),
2910(s), 2850(m), 161.000, 1490(s), 1450(s), 1070(m), 1030(m), 56 0(m);
NMR (CDC13, TM S) 16H multiplet at 0.72-3.48 ppm, 20H multiplet at 6.70-
7.40 ppm; mass spectrum, m/e (rel. intensity) metastable 368.3 (M +-92,6),
211(8), 129(8), 117(7), 115(7) 91(100), 65(10).
Anal. Calcd for C3411360: C, 88.65; H, 7.88. Found: C, 88.55;
H, 7.80.
Preparation of Reagents
Solutions of lithium aluminum hydride were prepared by ref luxing
distilled diethyl ether over solid LiA1H4 (Alfa.Inorganics) for 20 hours
followed by filtration in the dry box through a fritted glass funnel
using dried Celite as a filter aid. The clear solution was standardized
for aluminum content by EDTA titration. Diethylmagnesium was prepared by
• the reaction of diethylmercury 16 with magnesium metal at 60-80°C and a
165
solution in diethyl ether was standardized by magnesium analysis. Lithium
hydride was prepared by hydrogenation of t-BuLi or n-BuLi at 4,000 psi
pressure for 24 hours at roan temperature. Sodium hydride was obtained
as a 50% mineral oil dispersion from Ventron Hydrides Division. Lithium
alkoxides were prepared by the reaction of alcohols with n-butyllithium
or activated LiH.
Preparation of MgH2 slurry in TIE
17
Lithiun aluminum hydride (20 mniole) in diethyl ether (32 ml) was
allowed to react with a diethyl ether (50 ml) solution of diethylmagnesium
[(C2H5 ) 2Mg]. (20 mmole) at room temperature with constant stirring. The
reaction mixture was stirred for about 1 hour and the insoluble solid
isolated by centrifuging the mixture and then removing the ether solution
using a syringe. The resulting solid was washed with freshly distilled
diethyl ether and finally a slurry was made by addition of freshly dis-
tilled THF.
Anal. Calcd for MgH2 : Mg:H = 1.00:2.00. Found: 1.00:2.02.
Preparation of Lithium Alkoxymagnesium Hydrides
A known amount of lithium alkoxide in Et 20/THF was prepared by
the reaction of the appropriate alcohol in THE with n-hexane for about
1 hour at room temperature. The lithium alkoxide was allowed to react
with a slurry of MgH 2 in THE at room temperature in a 1:1 molar ratio
and the reaction mixture stirred for 3-5 hours during which time the
MgH2 dissolved. Analyses of the resulting clear solutions are given in
Table 28.
General Reaction of Hydrides With Ketones
A It) ml Erlenmeyer flask equipped with a Teflon coated magnetic
stirring bar was dried in an oven and allowed to cool under nitrogen or
argon. The flask was then sealed with a rubber serum cap, connected by
166
167
means of a syringe needle to a nitrogen/argon filled manifold equipped
with a mineral oil filled bubbler. The ketone solutions with internal
standard (tetradecane for 4-t-butylcyclohexanone and camphor, hexadecane
for 3,3,5-trimethylcyclohexanone and dodecane for 2-methylcyclohexanone)
were syringed into the flask and the known concentration of hydride
reagent (solution or slurry) added to the flask at room temperature.
After the designated time, the reaction was slowly quenched with H 2O and
dried over anhydrous MgSO4 . A 10 ft. 5% Carbowax 20 M on Chromosorb W
column (130°C column temperature) was used to separate the products of
4- t-butylcyclohexanone, 3,3,5 -trimethylcyclohexanone and camphor. A 15
ft. 10% diglycerol on Chromosorb W column (80°C) was used to separate the
products of 2-methylcyclohexanone. The order of elution for each ketone
was the same: the ketone first, the axial or exo-alcohol second, and
the equatorial or endo-alcohol last.
Qualitative Rate Studies of the Reaction of Lithium Alkoxymagnegiun Hydrides with 4-t-butylcyclohexanone
The aforementioned conditions for the reduction were set up at
the desired temperature. One syringe with the desired amount of ketone
and another syringe with a saturated NH 4C1 solution were placed through
the rubber serum cap. The ketone was added under an argon atmosphere
and at the desired time the reaction was rapidly quenched. The products
were analyzed in the normal manner.
168
CHAPTER III
RESULTS AND DISCUSSION
The magnesium hydride used in these studies was prepared by the ,
reaction of (C2H5 ) 211g with LiA1H4 in diethyl ether (eq. 1). A slurry of
Et20 + LiA1H
4 MgH2
4- + LiAl(C2H5)2H2 (1)
MgH2 (prepared by this method) in THE was prepared by removing the super-
natant solution containing the diethyl ether soluble LiAl(C 2H5 ) 2H2 by means
of a syrings and then adding freshly distilled THE to the resulting diethyl
ether-wet solid (MgH2 ) several times followed by solvent removal via
syrings in order to remove the last traces of LiAl(C 2H5 ) 2H2 .
Magnesium hydride prepared, as described above, was allowed to react
with lithium alkoxides (prepared according to equations 2 and 3) in equal
molar ratio in THE in order to prepare the desired lithium alkoxymagnesium
hydrides (equation 4, Table 28). :Lithium dialkoxymagnesium hydrides were
prepared according to equation 5 by allowing freshly prepared activated
LiH or NaH to react with the appropriate magnesium alkoxide. The magnesium
alkoxides were prepared by the reaction of Et 2Mg with freshly distilled
alcohol (equation 6).
n -BuLi ROH LiOR n-BuH T (2)
LiH ROB LiOR 14 (3)
169
LiMgH2(OR) LiOR MgH2
Mg(OR) 2 ----LiMgH(OR) 2
[NaMgH (OR) 2 ]
LiH (NaH)
Et2mg 2 HOR EtHis + Mg(OR) 2 (6)
Although MgH2 is insoluble in TIT, a clear solution results in most
cases when the lithium alkoxide is allowed to react with the MgH2 slurry.
It is at least theortically possible that the reaction of lithium alkoxide
with MgH2 proceeds to form THE soluble alkoxymagnesium hydrides (HMgOR) 5
according to equation 7. However, if this was the case, insoluble LiH
LiOR + MgH2 ----> LiH EMgOR (7)
would be formed and yet the reaction gives a clear solution. Thus the
possibility of forming HMgOR coupounds (eq. 7) does not seem likely.
However, when the solvent of the reaction described by eq. 4 was removed
under vacuum and the resulting solid analyzed by x-ray powder diffrac-
t18,19
ion, lines due to magnesium alkoxides were observed suggesting dispro-
portionation (eq. 8) when the solvent is removed. In solution, however, the
2 LiMgH2 (OR) -4 Li2.MgH4 Mg(OR) 2 (8)
integrity of LiMgH2 (OR) compounds has been confirmed by infrared (Figure 1)
and nmr spectral analysis (Figure 2) as well as molecular association stu-
170
dies reported earlier. 18 ' 19 Figure 1 shows that in the 1450 cm 1 region a
distinct absorption occurs for the LiOR compounds whereas it does not
appear for the LiMgH2 (0R), LiMgE(OR) 2 or HMgOR compounds when R = 2,2,6,6-
tetramethylcyclohexyl. However, a shoulder appears for HMgOR and LiMgH(OR) 2
conpounds that does not appear for LiMgH2(OR) compounds in the 1300 cm
-1
region. These similarities between HMgOR and LiMgH(OR) 2compounds could
suggest that disproportionation takes place according to equation 9.
LiMgH(OR) 2 HMgOR LiOR (9)
However the nmr spectra (Figure 2) show great differences for HMgOR and
LiMgH(OR)2 as well as for LiOR and LiMgH 2 (OR) compounds in the 1.0 ppm
region. A broad absorbance is observed for the HMgOR compounds shown
in Figure 2, however, for LiMgH(OR)2 compounds, a much cleaner spectrum
is observed in addition to a change in the chemical shift of the two
singlets assigned to the methyl groups on the cyclohexyl ring. HMgOR
compounds show two singlets at 0.98 and 0.93 ppm, but the singlets
associated with the LiMgH(OR) 2 compounds appear at 0.93 and 0.88 ppm.
Also, the LiOR coupounds have only one singlet at 1.00 ppm, but as noted
before, LiMgH(OR) 2 and EMgOR compounds both show two singlets. The NMR
were all conducted at the same concentration of reagent in THF-d 8 with
Me4Si as the internal standard with the same resolution. These observations
in addition to the molecular association studies reported earlier from this
laboratory18
show the separate identity of these compounds. The molecular
weight data show that the :HMgOR, LiMgH2 (OR) and LiMgH(OR) 2 compounds are
19 dimeric in TIE. Also it is important to note that the molecular weight
171
results did not indicate the presence of a mixture of LiOR and HMgOR
compounds when the LiMgH2 (0R) and LiMgH(OR) 2 compounds were being studied.
The LiMgH2 (0R) compounds prepared by the above methods were allowed
to react with four representative ketones, i.e., 4-t-butylcyclohexanone (I),
3,3,5-trimethylcyclohexanone (II), 2-methylcyclohexanone (III) and camphor
(IV). The purpose of these studies was to evaluate these new hydride
reagents as stereoselective reducing agents. We have compared the stereo-
chemical results with that of Lik1H4 which is considered to be the least
sterically hindered hydride that reduces cyclic and bicyclic ketones. For
example, Lik1H4 results in 10, 80, 24 and 9% equatorial or exo attack
respectively on ketones I, II, III and IV. On the other hand, MgH 2 reduces
ketones I, II, III and IV such that 23, 85, 35 and 8% equatorial or exo
attack, respectively, is observed.
Recently, both alkoxymagnesium hydrides (11MgOR)20
and dialkylamino-
magnesium hydrides (HMgNR2 ) 21 were prepared in this laboratory. The most
selective reagent among the HMgOR compounds studied was 2,6-diisopropyl-
phenoxymagnesium hydride which reduced ketones I, II, III and IV to give
83, 99, 99 and 98/ equatorial or endo attack, respectively.22
On the
other hand, trimethylsilyl-t-butylaminomagnesium hydride was the most
selective HMgNR2 compound studied. This hydride reduced ketones I, II,
III and IV by 73, 99, 98 and 95% equatorial or endo attack, respectively.23
With these data in hand, the LiMgH2 (0R) compounds were evaluated on a
comparative basis. According to Tables 29-32, lithium dihydrido-2,2,6,6-
tetramethylcyclohexoxymagnesiate and lithium dihydrido-2,2,6,6-tetrabenzyl-
cyclohexoxymagnesiate were the most selective in the reduction of ketone
(I) (Table 29) providing 89 and 86%, respectively, (Exps. 24 and 25) of the
axial alcohol. When allowed to react with ketone (II) (Table 30), the
tetrabenzyl derivative (Exp. 38) produced entirely the axial alcohol.
The tetramethyl- and 2-methylcyclohexoxy (Exp. 34) derivatives as well as
the neopentyl derivative (Exp. 30) produced 99% of the axial alcohol
when allowed to react with the ketone under the same conditions. Both the
tetramethyl- and tetrabenzyl- derivatives (Exp. 49 and 50) reduced ketone
(III) (Table 31) entirely to the axial alcohol. The neopentyl reagent
produced 99% (Exp. 43) of the axial alcohol. When ketone (IV) (Table 32)
was reduced, the tetrabenzyl reagent (Exp. 63) again was the most selective
providing 99% of the exo alcohol. The next best stereoselective reagent
was the tetramethyl reagent (Exp. 62) which produced 97% of the exo-alcohol.
From the data in Table 29, it can be seen that the amount of axial
alcohol increases as the steric bulk of the R group increases, particularly
when one compares the unsubstituted cyclohexyl derivative (70%) to the
2- ►ethyl derivative (78%), to the tetramethyl derivative (89%). There
also is a steady increase in the production of the axial alcohol when
proceeding from the primary alkoxy group (0CH 3 , 60%) to the secondary
alkoxy groups (-0Pr i and cyclohexoxy, 70%). As larger secondary groups
are introduced into the system (Exps. 17, neopentyl and 18, benzhydryl)
the amount of axial alcohol increased even further to 76%. These results
support the theory of steric approach control that predicts greater
approach of the reagent from the least hindered side of the substrate as
the steric requirement of the reagent increases. It is interesting to
note that the phenoxy reagents produced less axial alcohol as the steric
bulk increased (Exps. 20, 22 and 23). The unsubstituted phenoxy reagent
reduced ketone (I) with , almost the same stereoselectivity as MgH2 (Exp. 13).
172
Li+
41=
H
Mg
H Mg \
OR
RO H 2- ,/
2 Li
0.1•11. •■•
(A) (B)
0H
Mg ;Pr N_
H s
173
As diisopropyl and di-t-butyl groups were introduced in the phenoxy system
the amount of axial alcohol increased to only 50 and 59%, respectively.
Another observation along the same line shown the-tertiary derivative
(Exp. 16) is not as selective as expected. One might assume from these
observations that a small equilibrium amount of MgH2 is formed by dispro-
portionation (eq. 10) and that a significant portion of the reduction takes
LiMgH2
(OR) mg H2 + LiOR (10)
place with MgH2 . This is not unreasonable since the large steric bulk of
the alkoxy group should cause the hydride reagent to react very slowly with
the ketones studied thus allowing sufficient time for a competing side
reaction (eq. 10) to become significant.
Another explanation of the stereochemical results is that compounds
such as LiMgH2 (0- ) exist as a monomer (A) and LiMgH2 (0Pr 1) exists as
a diner (B). This is a resonable suggestion based on past findings24
that
very large alkoxy groups are poor bridging groups compared to smaller
alkoxy groups. This suggestion has been confirmed by molecular association
studies.18 In a monomer such as (A) it is clear that attack by a ketone
174
is less sterically hindered than on the dimer (B). It is likely that
reduction takes place by displacement of the solvent (S) from (A)
which should take place easier than the breaking of the Mg-}1 bond of
the dimer (B) in order to produce a coordination site.
When these same lithium alkoxymagnesium hydrides were allowed t o
react with ketone (II), 3,3,5-trimethylcyclohexanone (Table 30), the tetra-
benzyl derivative (Exp. 38) produced only the axial alcohol. Again the
phenoxy and the diisopropylphenoxy (Exps. 33 and 35) derivatives pro
duced the axial alcohol in almost the same yield as did MgH 2 40% versus
85% (Exps. 26 and 33). The di-t -butylphenoxy derivative (Exp. 36) also
produced a lesser amount (93%) of the axial alcohol than expected. These
results can also be explained by assuming disproportionation of the
reagent to MgH2 as suggested earlier or by assuming differences in
molecular aggragation of the reagents. The other derivatives produced
similar results with respect to each other and therefore no trends are
readily recognizable.
The reduction of 2-methylcyclohexanone (Table 31) with LiMgH 2 (0R)
compounds produced the same trends described earlier for 4-t-butylcyclo-
hexanone (Table 29). We once again observed the same low yield of axial
alcohol for the unsubstituted phenoxy derivative (Exp. 46, 38%) compared
to MgH2 (Exp. 39, 35%). The other phenoxy derivatives (Exps. 47, 66% and
48, 67%) as well as the tertiary butyl derivative (Exp. 42, 67%)
produced less axial alcohol than expected which is presumably due to
reduction of the ketone by a small equilibrium amount of MgH 2 formed
through disproportionation of the reagent (eq. 10) which would lower the
stereoselectivity of the reagent. Again, there is an increase in selec-
175
tivity when proceeding from the primary methoxy group (Exp. 40, 71%) .
to the secondary isopropoxy (Exp. 41, 96%) or cyclohexoxy (Exp. 45, 96%)
group of the reagent. As noted before, increasing the steric bulk in the
primary alkoxy reagents, i.e. neopentoxy (Exp. 43), or for the secondary
alkoxy reagents i.e. benzhydryloxy (Exp. 44) also increased the production
of the axial alcohol to 96%.
The reactions of these lithium alkoxymagnesium hydrides with
camphor (Table 32) produced similar trends as noted for the previous
reaction except for the dramatic differences observed for the methoxy
(Exp. 52, 12%), t-butoxy (Exp. 54, 10%), neopentoxy (Exp. 55, 9%) and
cyclohexoxy (Exp. 57, 15%) reagents. These reagents produced amounts of
endo-alcohol greater than MgH2 (Exp. 51, 8%) alone. This observation
implies that the above hydrides have less steric requirement when
approaching camphor than MgH2 . This is entirely possible. Other than
the observations just mentioned the trends described earlier seem to be
followed for camphor as well. An increase in stereoselectivity is
observed as the steric requirement of the reagent increases: cyclo-
hexoxy (Exp. 57, 15% endo-alcohol) < 2-methylcyclohexoxy (Exp. 59, 7%
endo-alcohol) < 2,2,6,6-tetramethylcyclohexoxy (Exp. 62, 3% endo-alcohol)
< 2,2,6,6-tetrabenzylcyclohexoxy (Exp. 63, 1% endo-alcohol). The
phenoxy derivatives (Exps. 58, 8%; 60, 8%; 61, 7% endo-alcohol) provided
the same selectivity as MgH 2 (8% endo alcohol) for presumably the same
reasons described earlier.
Another interesting observation is that the cyclohexoxy reagent
shows the lowest yield of alcohols (Exps. 19, 45, and 57) for the reduction
of ketones . I, III and IVbut not in the case of ketone (II) (Exp. 32).
176
Evidently enolization should be much greater in the case of ketone I,
III and TV because of the greater steric hindrance in the case of ketone
(II). Why enolization is greater for the cyclohexoxy derivative compared
to the other hydrides is not clear. It should also be noted that all of
the phenoxy derivatives provide the highest yields even though the ster-
eoselectivity is less than the other hydrides. In general, it appears
that all of the reagents capable of producing small amounts of MgH2 , as
described above, produced the least amount of enolization. Magnesium
,hydride itself produced a 100% 5 yield of the reduction products with all
of the ketones studied except for ketone (II) (Exp. 26). Ketone (II) is
the ketone which the cyclohexoxy reagent .did not enolize the ketone (Exp.
32) at all. However, it should also be noted that for the reduction of
ketone (II) another secondary alkoxy reagent (benzhydryloxy) did have the
most enolization associated with it (Exp. 31). The reason why MgH 2 or
very bulky reagents lower the amount of enolization is probably due to
steric reasons.
It was also desired to determine if the degree of stereoselectivity
was higher in the initial stages of the reaction than after equilibrium
had been reached. In this connection, 4-t--butylcyclohexanone was allowed
to react with a 100% excess of 2,2,6,6-tetrabenzoxydihydridomagnesiate.
The results are give in Table 33 and Figure 3. As can be seen from the
data, the lower the temperature, the greater the observed enolization
to reduction ratio. At -25°c, for example, the major product (64%) after
quenching is the starting ketone, 4-t-butylcyclohexanone. It should also
be noted that the initial reaction is very fast and that after 5 hours no
more than a 5% change was observed in the ratio of alcohols. The thermo-
177
dynamic product is the equatorial alcohol, and if 4-t-butylcyclohexanone
is allowed to react under equilibrium conditions inherent for Meerwein-
Ponndorf-Verley or Birch reductions, the equatorial alcohol is produced
in 98-99% yield.25 For lithium alkoxymagnesium hydride reductions, never
less than 86% of the axial alcohol is observed. It appears that most of
the stereochemical composition always takes place during the time between
60 and 18,000 seconds indicating some, but slow equilibrium.
Tables 34-37 list the results of the reactions of ketones (I), (II),
(III) and (IV) with lithium dialkoxymagnesium hydrides prepared according
to equation 5. The most selective reagents were the bis - tetramethyl- and
bis-tetrabenzylcyclohexoxy hydrides (Exps. 67a and 68a) which reduced
ketone (I) (Table 34) to provide 89 and 85% axial alcohol, respectively.
However a large amount of enolization accompanied the reaction (66 and
60%, respectively). On the other hand, Exp. 65a shows that the bis-di-t-
butyl derivative enolized only 18% of the ketone while reducing the ke-
tone to 81% of the axial alcohol. The sodium reagent (Exp. 66a) not only
produced a 55:45 axial to equatorial alcohol ratio, but also enolized 85%
of the ketone. When LiH and LiOR were allowed to react under similar
conditions (Exp. 64a) a 74:26 ratio of axial to equatorial alcohol was
observed but 70% of the ketone was enolized.
When ketones (II), (III) and (IV) (Tables 35-37) were allowed to
react with these reagents, lesser amounts of enolization were observed
with very stereoselective results. All the reagents studied produced 99%
to 100% axial alcohol when allowed to react with ketones (II) and (III)
(Tables 35 and 36, respectively). The reactions with camphor (ketone IV,
Table 37) produced greater than 90% exo-alcohol with little enolization
178
except for Exps. 64d and 66d which produced 65 and 62% respectively of the
starting ketone. These reagents represent a method of using lithium and
sodium hydride for reduction which has not been previously reported, and
further development is now under way.
CHAPTER IV
CONCLUSION
A series of lithium alkoxyinagnesium hydrides, LiMgH2 (0R), were
prepared and allowed to reduce 4-t-butylcyclohexanone (I), 3,3,5-trimethyl-
cyclohexanone (II), 2-methylcyclohexanone (III) and camphor (IV). It was
found that very bulky secondary cyclic alkoxy groups such as 2,2,6,6-tetra-
methyl- or benzylcyclohexoxy were very stereoselective in the reduction
of these ketones. For example, LiMgH 2 ( ) reduced ketone (I) to pro-
vide 89% of the axial alcohol compared to BMg which provided 83%
of the axial alcohol.
The LiMgH(OR) 2 when R = ) or Oreagents were also found to
reduce ketones (I), (II), (III) and (IV) stereoselectively but to a lesser
extent and with more enolization than observed for the LiMgH2 (OR) reagents.
179
Product
LiMgH2 (OCH.3 )
LiMgH2(0Pr i )
LiagH2 (0But )
LiMgH2 (OCH2Bu )
LiMgH2 (OCHPh2 )
LiMgH2 (0-11))
LiMgH2(OPh)
LiMgH2 (0-10
LiMgH2 (
LiMgH (
LiMgH
LiMgH
Table 28. Preparation of Lithium . Alkoxymagnesium Hydrides [LitigH2 (0R)] by the Reaction of Magnesium Hydride with Lithium Alkoxides in a 1:1 Ratio.
180
Exp. Reaction Solubility Time (hr) in THE Li
Analysis (Ratio) : Mg : H : ROH
1 10 Insoluble 1.05 1.00 1.91 1.02
2 7 Soluble 1.04 1.00 1.93 1.01
3 7 Soluble 1.04 1.00 1.93 1.03
4 6 Soluble 1.03 1.00 1.95 1.02
5 8 Soluble 1.04 1.00. 1.92 1.02
6 7 Soluble 1.00 1.02 1.95 1.03
7 8 Insoluble 1.03 1.02 1.90 1.04
8 5 Soluble 1.02 1.00 1.96 1.00
9 7 Soluble 1.03 1.00 1.92 1.03
10 6 Soluble 1.02 1.02 1.94 1.00
11 5 Soluble 1.01 1.01 1.96 1.00
12 5 Soluble 1.01 1.01 1.97 1.00
13
MgH2
14
LiMgH2 (0CH3)
15
LiMgH2 (0Pri )
16 LiMgH2 (OBut )
17
LiMgH2 (OCH2But)
18
LiMgH2 (OCHPh2 )
19 LiMgH2 (0-11D)
20
LiMgH (
21
LiMgH2 (011))
22
LiMgH2 (
23
LiMgH (
24 L iMg H
25 LiM H,(0
181
Table 29. Reaction of 4=tBdtylcyclohexanone with LitigH 2 (0R) Compounds at Room Temperature in TIT Solvent.
Relative Yield (%) d Exp. Hydride Axial-OH Equatorial-OH Yield(e.
23 77 100
60 40 95
70 30 91
59 41 98
76 24 95
76 24 90
70 30 52
28 72 100
78 22 83
50 50 100
59 41 99
89 11 86
86 14 85
a) Normalized % axial alcohol + % equatorial alcohol = 100%.
b) Yield was determined by glc and based on an internal standard.
Exp. Hydride
26 MgH2
27 LiMgH2 (OCH )
28 LiMgH (0Pri )
29 LiMgH2 (0But )
30 LiMgH2 (OCH2But )
31 LiMgH2(OCHPh
2)
32 LiMgH2 (0-C) )
33 LiMgH (
34 LiMgH2 (0-2)
35 LiMgH2 (0
36 LiMgH (
37 LiMgH (
Table 30. Reactions of 3,3,57Trimethylcyclohexanone with LiMgH 2 (0R) Compounds at RoOM Temperature in THE i2 Molar Ratio.
182
Relative Yield Axial-OH Equatorial-OH Yieldb
85 15 92
95 5 93
98 2 84
96 5 100
99 1 88
97 3 80
96 4 100
90 10 100
99 1 100
90 10 100
93 7 100
99 1 100
38 LiMgH2 0 41) 100 0 96
a) Normalized % axial alcohol + % equatorial alcohol = 100%. b) Yield was,determined by glc using an internal standard.
39
40
41
42
43
44
45
46
47
48
49
50
MgH2
LiMgH2 (OCH3)
LiMgH2 ( r )
LiMgH2 (OBut)
LiMgH2 (OCH2But )
LiMgH2 (OCHPh2 )
LiMgH2 (0-10 )
LiMgH (
LiMgH (
LiMgH (
LiMgH
LiMgH (
4
183
Table 31. Reactions of 2-Methylcyclohexanone with LiMgH2 (0R) Compounds. at Room Temperature in THE Solvent in 1:2 Ratio.
Relative Yield (%)9- Exp. Hydride Axial-OH Equatorial-OH Yield(%)b
35 65 100
71 29 .94
96 4 92
67 33 100
99 1 88
98 2 88
96. 78
38 62 100
66 34
67 •33 95
100 0 87
1 00 85
a) Normalized % axial alcohol + % equatorial alcohol = 100%.
b) Yield was determined by glc using an internal standard.
Exp. Hydride
51
MgH2
52
LiMgH2 (OCH3 )
53
LiMgH2 (0Pri )
54
LiMgH2( ut)
55
LiMgH2 (OCH2But
)
56
LiMgH2 (OCHPh2 )
57
LiMgH (
58
LiMgH2 (
59
LiMgH2 (
60
LiMgH2
61
LiMgH2
62
LiMgH2
Table 32. Reactions of chor with LitIgH2 (0R) dbMpounds at Room Temperature in THE Solvent in 1:2 Molar Ratio.
184
Relative Yield (%) a Endo-OH. Exo-OH Yield M b '
8 92 100
12 88 88
8 92 100
10 90 91
9 91 100
4 96 100
15 85 52
92 100
7 93 83
92 100
7 93 100
3 97 95
1 99 95
a) Normalized %,axial alcohol + % equatorial alcohol = 100%.
b) - Yield was determined by glc using an internal standard.
185
Table 33. The Reaction of L24H2 ( ) in THE Solvent with 4-t-Butyl- -
Cyclohexanone at Various Temperatures and Reaction Times in 2:1 Molar Ratio.
2 75 5 67 10 56 15 47 30 37 45 37
18,000 35
Axiala Alcohol (%)
Equatoriala Alcohol (%)
91 9 90 10 90 10 90 10 90 10 86 14
90 10 91 9 90 10 90 10 89 11 88 12 86 14
90 10 90 10 91 9 92 8 91 9 89 11 89 11 86 14
RT
2 76 5 65 10 47 15 38 30 22 45 15 60 15
18,000 15
TIME Recovered TEMP.0C (sec) Ketone (%)
-25° 2 75 5 70 10 67 15 67 30 66
18,000 64
a) Normalized % axial alcohol ± % equatorial alcohol = 100%..
Recovered Ketone (%)
. Reduction Products Axial - Alcohol •(%) a. Equatorial•Alcohol•(%) a Yield
70 74 26 28
18 81 19 80
85 55 45 10
66 89 11 30
60 85 15 37
Exp. Reagent
. 64a LiH + Li'
65a LiH + Mg() 2 X
66a NaH + Mg(
67a LiH + Mg(
68a LiH + Mg(
Table 34. Reactions of 4-t-Butylcyclohexanone with Metal Hydrides and Magnesium Alkoxide at Room Temperature in THE Solvent and in 1:2 Molar Ratio for 24 Hours.
a) Normalized % axial alcohol + % equatorial alcohol = 100%.
b) Yield was determined by glc using an internal standard.
64b LiH + Li
65b LiH + Mg(
66b NaH + Mg(
67b LiH + Mg(
14
99
_ -
30 99
35
99
a) Normalized % axial alcohol + % equatorial alcohol = 100%.
Table 35. Reactions of 2-Methylcyclohexanone with Metal Hydrides and Magnesium Alkoxides at Room Temperature in THE Solvent and in 1:2 Molar Ratio for 24 Hours.
Exp. ' Reagent Recovered Reduction Products Ketone (%) Axial Alcohol (%)
a Equatorial Alcohol (%) a Yield (%)
b
b) Yield was determined by glc using an internal standard.
1 85 .
1 65
1 60
Table 36. Reactions of 3,3,5-Trimethylcyclohexanone with Metal Hydrides and Magnesium Alkoxides at Room Temperature in THE Solvent and in 1:2 Molar Ratio for 24 Hours.
Exp. Recovered Reduction Products
Reagent Ketone (%) Axial Alcohol (%) a Equatorial Alcohol (%) a Yield (%'
64c LiH + Li
65c LiH + Mg(
66c NaH + Mg(
67c LiH + Mg(
68c LiH + Mg(
8 100 0 90
35 100 0 60
42 100 0 55
a) Normalized % axial alcohol + % equatorial alcohol = 100%.
b) Yield was determined by glc using an internal standard.
Exp. Reagent
64d LiH + Li
65d LiH + Mg(
66d NaH + mg(
67d LiH + Mg(
68d LiH + Mg(
Table 37. Reactions of Camphor with Metal Hydrides and Magnesium Alkoxides at Room Temperature in THE Solvent and in 1:2 Molar Ratio for 24 Hours.
Recovered Ketone (%) Axial Alcohol
Reduction Products (%)a Equatorial Alcohol (%)a Yield (%) b
65 91 9 30
0 95 5 100
62 90 10 36
11 97 3 87
16 96 4 81
a) Normalized % axial alcohol + % equatorial alcohol = 100%.
b) Yield was determined by glc using an internal standard.
a) P/Ig0-.
LiMgH2 (0
c) LiKgH(0
d) Li0
FIGURE 2
NMR Spectra of Simple and Complex
Metal Alkoxides
192
FIGURE 3
The Reaction of timgH2 (
In THE with 4-t-butylcyCIohexanone in 2:1 Ratio
o -25°C
A 0°C
. RT (25 °C)
194
LITERATURE CITED
1. E. C. Ashby, J. J. Lin and A. B. Goel, J. Org. Chen., 43, 1557(1978).
2. H. 0. House, "Modern Synthetic Organic Reactions", W. A. Benjamin, New York, N. Y., 1972, p. 44ff.
3. S. Krishnanurthy and H. C. Brown, J. Au. Chem. Soc., 98, 3383(1976).
4. E. C. Ashby and J. R. Boone, J. Org. Chem., 41, 2890(1976).
5. E. C. Ashby and A. B. Goel, Inorg. Chem., (in press).
6. E. C. Ashby and R. G. Beach, Inorg. Chem., 10, 906(1971).
7. E. C. Ashby, J. J. Lin and A. B. Goel J. Org. Chem., 43 156 1 (1978).
8. E. C. Ashby and A. B. Goel, Inorg. Chem., (in press).
9. D. F. Shriver, "The Manipulation of Air-Sensitive Compounds", McGraw-Hill, New York, N. Y., 1969.
10. E. C. Ashby and R. D. Schwartz, J. Chem. Educ., 51, 65(1974).
11. H. Gilman and A. H. Haubein, J. Am. Chem. Soc., 66, 1515(1944).
12. S. Borg, M. Fetizon, P. Laszlo and D. H. Williams, Bull. Soc. Chim. France, 2541(1965).
13. P. Granger and M. M. Claudon, Bull.Soc. Chim. France, 753(1966).
14. M. Boyer, M. M. Claudon, J. Lemaire and C. Bergamini, Bull. Soc. Chin. France, 1139(1966).
15. M. Boyer, M. M. Claudon, J. Lemaire and C. Bergamini, Bull. Soc. Chin. France, 2152(1964).
16. E. C. Ashby and R. C. Arnott, J. Organometal. Chem., 14, 1(1968).
17. E. C. Ashby and R. G. Beach, Inorg. Chem., 2, 2300(1970).
18. E. C. Ashby and A. B. Goel, Inorg. Chem., (in press).
19. Dr. A. B. Goel of this laboratory conducted the molecular associa-tion and x-ray powder diffraction studies.
20. E. C. Ashby, J. J. Lin and A. B. Goel, Inorg. Chem., (in press).
196
197
21. E. C. Ashby, J. J. Lin and A. B. Goel, J. Org. Chem., 43, 1564(1978).
22. E. C. Ashby, J. J. Lin and A. B. Goel, J. Org. Chem., (in press).
23. The reagent bis-dicyclohexylaminamagnesium hydride was prepared and allowed to react under similar conditions an it was observed that the amount of axial alcohol was increased from 73% for the trimethyl-silyl-t-butylaminomagnesium hydride to 90%. This raises the possi-bility of using more hindered cyclohexyl substitutents in order to increase the selectivity of these reagents.
24. E. C. Ashby and G. E. Parris, J. Am. Chem. Soc., 93, 1206(1971).
25. a) E. L. Eliel, R. J. L. Martin and D. Nasipuri, Org. Syn., 47, 16 (1967). b) E. L. Eliel, Rec. Chen. Progr., 22, 129(1961). c) J. C. Richer and E. L. Eliel, J. Org. Chem., 26, 972(1961). d) E. L. Eliel and D. Nasipuri, J. Ora. Chem., 30, 3809(1965). e) J. W. Huffman and J. T. Charles, J. Am. Chem. Soc., 90, 6486(1968).
PART IV
CONCERNING SALT EFFECTS ON THE STEREO SELECTIVITY OF
ORGANOMETALLIC COMPOUND ADDITION TO KETONES
198
CHAPTER I
INTRODUCTION
Background
Recently, it was reported that a mixture of CH3Li and LiCu(CH 3 ) 2
provides unusually high stereoselectivity (94% equatorial attack) in
the methylation of 4-tert-butylcyclohexanone compared to reaction of
CH3Li or LiCu(CH
3 ) 2 alone.
1 It was suggested that "a bulky, highly
reactive cuprate having the stoichiometry Li 2Cu(CH 3 ) 3 or Li3Cu(CH
3)4"
was formed when CH 3Li and LiCu(CH ) 2 are allowed to react; and, re-
action of these cuprates with the ketone would explain the observed
results. However, molecular weight measurements indicate that 2 3
Li2Cu(CH3 ) 3 is monomeric in diethyl ether and THF, whereas CH 3Li is
4 tetrameric and LiCu(CH
3)2 is dimeric. As a monomer, Li
2Cu(CH 3 )
3
should not be considered more bulky than a tetrametric molecule such
as CH3Li. Reactions of Li2Cu(CH3)3' LiCu(CH
3)2 and LiCu
2(CH
3)3
in
both diethyl ether and THF with selected, enones indicates that
Li2Cu(CH 3 ) 3 i s - only slightly more reactive than LiCu(CH 3 ) 2 toward
conjugate addition.5
Therefore, the hypothesis that Li 2Cu(CH3 ) 3 , when
present in a mixture of CH3Li and LiCu(CH 3 ) 2 in diethyl ether, is a
"bulky, highly reactive cuprate" is questionable.
The CH3Li-LiCu(CH3 ) 2
mixture used to methylate 4-tert-butyl- \
cyclohexanone was prepared by reacting CH 3Li with CuI in a 8:3 molar
ratio in diethyl ether solvent. In such a mixture at least three
species are present: LiCu(CH 3 ) 2 , CH3Li and LiI. The reaction of-any
199
one of these compounds with 4-test-butylcycloheXanbne fails to produce
the unusual stereochemistry reported above. One can suggest four
possible explanations for this steteoselectivity: (1) CH 3Li.reacts
with LiCu(CH3
)2 to fort a complex which then reacts with the ketone;
1
(2) CH3Li reacts with LiI to form a complex (a reaction known to pro-
2 duce Li4 (CH 3 ) 3I /which then reacts with the ketone; (3) LiCu(CH3 ) 2 and
Lii react to form a complex which then reacts with the ketone; (4) one
of the species in solution reacts with the ketone to form a complex
followed by reaction of the complexed carbonyl compound with CH3Li.
1 6 Recently, low temperature H NMR evidence was reported for the
existence of Li2Cu(CH
3)3 in a mixture of CH3Li and LiCu(CH )
2 in di-
methyl ether, tetrahydrofuran and diethyl ether solvents. No evidence
was found to indicate the presence of any higher order complexes, such
CH3Li + LiCu (CH3 2 ) Li2 Cu (CH 3 ) 3 3
as Li 3 Cu(CH3
4 . 3 ) The reaction CHLi-LiCu(CH
3)2 with 4-test-butyl-
cyclohexanone in THF did not yield any increased stereoselectivity when
compared to CH3Li alone. Since Ashby, et al. have determined that
Li 2Cu(CH 3 ) 3 exists in both ether and THF and is monomeric in both sol-
2 vents, it is doubtful that Li 2Cu(CH 3 ) 3 would react with 47tert-butyl-
cyclohexanone in diethyl ether to give unusual stereoselectivity when in
THF no trace of unusual stereoselectivity is observed. Therefore, one
is led to question that the observed stereoselectivity in diethyl ether
is due to the reaction of Li 2Cu(CH3 ) 3 with the ketone.
The stereochemical improvement in the CH Li-LiCu(CH 3 ) 2 reagent
200
201.
in diethyl ether cannot be explained by assuming that a complex between .
CH 3Li and Lii (formed in the reaction of CH3Li with CuI) is reacting
with the ketone. A mixture of CH3Li and Lii or LiBr (Table 11) while
giving some improvement in stereoselectivity, does not give the selec-
tivity observed with the CH 3Li-LiCu(CH 3 ) 2 mixture, Also ., a mixture of
CH3Li and Lii or LiBr in THF gives no improvement in stereoselectivity
over CH3Li alone. It is known that CH3Li forms complexes with both
7 .
Lii and LiBr in THF. Likewise, the stereochemical improvement
cannot be explained by assuming that a complex between either LiCu(CH 3 ) 2
or Li2 Cu(CH 3 ) 3 and LiI is reacting with the ketone, since a halide free
mixture of CH3Li and LiCu(CH3) 2 gives the same high stereoselec-
tivity
The only reasonable possibility remaining then is tnat um3L1
reacts with a complex between LiCu(CH3 ) 2 and ketone (eq. 2). This
LiCu(CH3 ) 2 + R2C (2) ' .L (CH3 ) 2
suggestion also explains why there is no rate enhancement or increase in
stereoselectivity when the same reaction is conducted in THF; i.e., the
ketone would not be expected to compete effectively with THF solvent
molecules for coordination sites on lithium. The unusual rate enhancement
in diethyl ether is explained on the basis that the concentration of
ketone complexed to LiCu(CH3 ) 2 , Lii, et. would be expected to be
considerably higher in ether than in THF and certainly the complexed
carbonyl compound would be much more reactive than the uncomplexed
13 carbonyl. Recent reports involving C NMR have confirmed that no
202
complex formation exists between CH3Li and lithium salts such as LiC104 , 7
but complex formation does take place between LiC104 and the carbonyl
group7 ' 8 in diethyl ether and yet as stated earlier, a dramatic increase
in stereoselectivity is observed with LiC104 just as in the case of
Liem(CH3 ) 2 .
Purpose
In order to complete a more detailed study of the unusal stereo-
selectivity found in the alkylation of cyclohexAnones with CH3Li in the
presence of lithium salts, several metal salts were added to the reaction
of CH3Li and 4-t-butylcyclohexanone. In an attempt to expand the scope
of the reaction, other ketones (e.g. 2-methyl- and 3,3,5-trimethylcyclo-
nexanone) and other organolithium reagents (e.g. t-butyl-and phenyllithium)
were also allowed to react in the presence of LiC10 4 . Similar studies
were also conducted with organomagnesium and aluminum reagents in place
of CH3Li.
CHAPTER II
EXPERIMENTAL
ApvAtatus-
Reactions were performed under nitrogen or argon at the bench using
Schlenk tube techniques or in a glove box equipped with a recirculating
system using manganese(II) oxide to remove oxygen and dry ice-acetone traps
to remove solvent vapors. 9 Calibrated syringes equipped with stainless
steel needles were used for transfer of reagents. Glassware and syringes
were flamed and cooled under a flow of nitrogen or argon. Ketones, metal
salts and solutions of internal standards were prepared by weighing the
reagent in a tared volumetric flask and diluting with the appropriate
solvent. GLPC analyses were carried out on an F and M Model 700 or Model
720 gas chromatograph. Flame photometry was conducted on a Coleman Model
21 Flame Photometer.
Analyses
Gas analyses were carried out by hydrolyzing samples with hydro-
chloric acid on a standard vacuum line equipped with a Toepler pump. 10
Magnesium was determined by titrating hydrolyzed samples with standard
EDTA solution at pH 10 using Eriochrome-Black T as an indicator.
Aluminum was determined by adding excess standard EDTA solution to hydro-
lyzed samples and then back titrating with standard zinc acetate solution
at pH 4 using dithizone as an indicator. Lithium reagents were analyzed
by the standard Gilman double titration method (titration of total base
203
204
followed by titration of residual base after reaction with benzyl
chloride)11 or by flame photometry. The amount of active C-Mg and
0-Li was determined by titrating the active reagent with dry 2-butanol
in xylene using 2,2'-diquinoline as an indicator.
Materials
Tetrahydrofuran (Fisher Certified Reagent Grade) was distilled
under nitrogen from NaA1H4 and diethyl ether (Fisher Reagent Grade) from
LiA1H4 prior to use. Methyllithium in THE or diethyl ether was prepared
by the reaction of (CH3 ) 2Mg with excess lithium metal dispersion (Alfa),
30% in petrolatum, which was washed repeatedly with ether/pentane until
clean under an argon atmosphere prior to use. Both solutions were stored
at -78°C until ready to use.
Dimethylmagnesium was prepared12
by the reaction of dimethylmercury
with magnesium metal (ROC/RIC) at 40-60°C in the absence of solvent. The
resulting (CH3 ) 2Mg was extracted from the gray solid with ether and the
resulting solution standardized by magnesium analysis.
Trimethylaluminum (Ethyl Corporation) was distilled under vacuum in
a glove box and standard solutions were prepared in diethyl ether and THE.
The resulting solutions were standardized by aluminum and methane analysis.
Lithcoa's t-butyllithium and n-butyllithium, MC/B methyllithium and
PCR Incorporated phenyllithium were analyzed prior to use for active C--Li
by the Watson and Eastman procedure described in the Analytical Section.
The methyllithium reagent was also analyzed by methane gas analysis using
standard vacuum line techniques. All reagents were hydrolyzed prior to use
and the organic fractions subjected to glpc analysis.
LiA1H4 (Alfa Inorganic) was suspended in refluxing ether or THE for
205
24 hours, then filtered in a glove box using a glass fritted funnel and
Celite filter aid. The clear solutions were standarized by aluminum and
gas analyses prior to use.
Alane, All3 , in THE or diethyl ether was prepared by the reaction
of 100% H2 SO4 with LiA1H4 in the appropriate solvent
13 at -78°C followed
by filtration of the resulting Li2 SO4 in the dry box. Analysis:
Li:Al:H = 0.02:1.00:3.00.
Camphor, norcamphor, 2-methylcyclohexanone, 3,3,5-trimethylcyclo-
hexanone and 4-t-butylcyclohexanone were obtained from Aldrich Chemical
Company or Chemical Samples Company and sublimed or distilled and stored
over 4A molecular sieves prior to . use.
Lithium salts: lithium perchlorate, lithium iodide and lithium
bromide were purchased from Alfa Inorganics and dried under vacuum at
100°C overnight.
Sodium tetraphenyl borate and tetramethylammonium iodide were
purchased from Fisher and used without further purification.
Lithium methoxide and t-butoxide were prepared by the reaction of
n-butyllithium in hexane with a stoichiometric amount of the appropriate
alcohol under anhydrous conditions using an argon atmosphere.
Magnesium methoxide and t-butoxide as well as bis-diisopropyl-
aminomagnesium were prepared in diethyl ether by the reaction of di-
methylmAgnesium with a stoichiometric amount of the appropriate alcohol
or amine under an argon atmosphere.
Aluminum methoxide and isopropoxide and tris-diisopropylamino-
aluminum were prepared in TIP by the reaction of alane with a stoichio-
metric amount of the appropriate alcohol or amine under an argon atmo-
sphere.
206
If the lithium, magnesium or aluminum salts were desired in another
solvent, the original solvent was removed by vacuum and replaced by the
desired solvent which was freshly distilled. This procedure was repeated
three times.
General Reactions of Ketones
A 10 ml Erlenmeyer flask with a teflon coated magnetic stirring
bar was dried in an oven and allowed to cool under argon or nitrogen.
The flask was then sealed with a rubber serum cap and connected by means
of a syringe needle to an argon or nitrogen filled manifold. The amount
of organometallic reagent (ca. 0.1-1.0 mmole) in THE or diethyl ether was
added to the flask using a syringe,and the addition of the metal salt
solution followed, if needed. The temperature was controlled by a dry
ice-acetone bath, then the calculated amount of ketone in THE or diethyl
ether was added to the stirred mixture.
The reverse addition (organametallic reagent added to a mixture of
ketone, internal standard and metal salt at -78 °C) produced the same
results. After the designated time, the reaction was quenched with the
slow addition of methanol and dried over anhydrous MgSO4 . A 12 ft. 20%
FFAP on Diatoport S column (column temperature: 150oC,elium flow rate:
60 ml/min) was used to separate the products of 4-t-butylcyclohexanone.
The retention time was 13.0 minutes for n-C 14 H30' 33.0 minutes for cis-1-
methyl-4-t-butylcyclohexanol, 38.0 minutes for 4-t-butylcyclohemanone and
41.5 minutes for trans-1-methyl-4-t-butylcyclohexanol. A 14 ft. 10%
diglycerol on Diatoport S column at 80°C was used to separate the products
of 2-methylcyclohexanone. The rentention time was 4.4 minutes for the
ketone, 5.2 minutes for the cis-alcohol, 9.5 minutes for the trans-alcohol
207
and 16.1 minutes for n-C 14 1130 . A 10 ft. column of 20% SUB on Chromosorb
W at 180°C and flow rate of 60 nil/minute gave the following retention
times for ketone, axial alcohol and equatorial alcohol for the methylation
of 3,3,5-trimethylcyclohexanone: 5.0, 4.3 and 6.0 minutes.
The isomeric alcohols resulting from the methylation of norcamphor
could not be separated by glpc. In addition, the isomeric alcohols
resulting from the phenylation of all ketones studied could not be
determined by glpc because of dehydration. Therefore, the isomer ratios
in these cases were determined by NMR analysis. Ashby, Laemmle and Roling 14
determined the methyl singlet for the resulting exo-alcohol from the methyl-
ation of norcamphor to be 73 Hz in benzene and the endo-alcohol methyl
group singlet to be 74 Hz. These results were confirmed by the author.
The identification of the phenylation products of all ketones was
determined by NMR spectroscopy utilizing the peak areas of the hydroxyl
protons of the alcohols in DMSO-d 6 . In these cases, work up of reaction
mixtures was carried out as follows. After the reaction was complete,
distilled water was added in order to effect hydrolysis. The solution.
was then transferred to a separatory funnel and washed several times
with distilled water. The ether layer was separated and allowed to
evaporate and DMSO-d 6 added to the sample. The sample was then dried
over Linde 4A molecular sieves and transferred to an NMR tube. This
procedure minimized dehydration and equilibration. Benzaldehyde was
added as the internal standard.
In the case of phenylation of 4-t-butylcyclohexanone, the chemical
shifts for the axial and equatorial hydroxyl protons are 4.56 ppm and
4.73 ppm respectively. - The assignment of each alcohol hydroxyl peak to a
208
particular isomer was baed on several reports in the literature con- ,
cerning their chemical shifts in DMSO and DMSO-d6. 15 The identification
of the phenylation products of 2•methylcyclohexanone was accomplished
according to the procedure of Luderer, Woodall and Pyle.16
The 1H NMR
showed the methyl doublet of the axial alcohol to be at 0.60 ppm and the
methyl doublet associated with the equatorial alcohol to be at 0.62 ppm.
Identification of the tbutylation products of 4-t-butylcyclohexanone
was accomplished according to the method of Meakins, et a1. 15c The axial
hydroxyl proton in DMSO-d 6 is 3.49 ppm. Finally, the identification
of the t-butylation products of 2-methylcyclohexanone was accomplished by
the method of Ficine and Maujean.17
209
CHAPTER III
RESULTS AND DISCUSSION
OtgdziOlithiunilteadtiOns
The detailed results of stereoselective methylation in diethyl
ether of 4-t-butyicyclohexanone (eq. 3) by main group metal organometallic
__f__ /:„:2::c113 11■
0 (3)
reagents in the presence of equimolar amounts of various metallic salts
[ca. LiBr, LiI, LiC104 , NaBPh4 , Me4NI, LiOBut, Mg(OPri) 2 , Al(OPr i) 3 and
Al(NPt) 31 are reported in Tables 38, 45 and 47.When CH3Li was allowed to
react with 4-t-butylcyclohexanone in diethyl ether solvent, 65% of the
axial alcohol was formed. However, when LiOBu t , LiBr,18 LiI18 r LiC104
18
(Table 38) was added to an ether solution of CH 3Li before it was added to
the ketone, the amount of axial alcohol formed increased to 75, 76, 81
or 92% respectively. When THE was substituted for diethyl ether, no
increase in stereoselectivity was observed. A number of other salts
(Exps. 2-7) produced no effect or only a slight increase ( 5%) in the
formation of axial alcohol. For the cases when Al(NPr2) 3' Al(OPr i)3'
Mg(OCH3 ) 2' Mg(OPri) 2 and LiOBu
t were present, the total product yield
decreased to 80, 85, 83, 85 and 73 % respectively. This compares to the
total product yield > 92% for the other reactions. This result is pre-
sumably due to enolization.
Because LiC104 had the greatest effect on the stereochemistry of
this reaction, it was decided to study the scope of this reaction using
LiC104 but varying the nature of the ketone and the organolithium reagent.
When ketones other than 4-t-butylcyclohexanone (Table 39) were
allowed to react with CH3Li in the presence of LiC10
4' only 2-methylcyclo-
hexanone showed a modest increase in the production of the axial alcohol
(88% without and 96% with LiC10 4 added). The other ketones, 3,3,5-tri-
methylcyclohexanone and norcamphor did not show any increase in the forma-
tion of the axial alcohol since only the axial or endo-alcohol was ob-
served in the absence of LiC104'
However, even these ketones showed a
pronounced rate enhancement when allowed to react with CH3Li in the
presence of LiC104 (Table 40). In the absence of LiC10
4' the reaction of
CH3Li with 4-t-butylcyclohexanone, 2-methylcyclohpwanone or norcamphor
was essentially over in one hour; however, when LiC104 was present, the
reaction was complete in less than 10 seconds at -78°C. The substantial
rate enhancement is probably due to the increased reactivity of the newly
formed complex between the carbonyl group and LiC104 (eqs. 4 and 5).
Equation 4 illustrates the proposed transition state for the
reaction of CH3Li with 4-t-butyIcyclohexanone which provides 65% of the
axial alcohol upon hydrolysis.
210
(4)
t/S
H;C..1 /
35%
If LiCIO4 is introduced (eq. 5), complexation between LiC10 4 and
A
‘C104
(5)
S = Solvent
211
212
the carbonyl group takes place increasing the rate of subsequent attack by
CH3Li and also influencing the subsequent stereochemistry of the reaction.
It is not surprising that complexation between LiC1 04 and the carbonyl
group increases the rate of reaction with CH3Li since the complexed
carbonyl group should be highly polarized as in other cases of acid
catalysis of carbonyl compounds. This complexation also accounts for the
increase in the formation of the axial alcohol because complex A does not
have the 3,5-diaxial hydrogen interaction as illustrated in complex B.
The complexation also explains the results obtained when THE is the
solvent: namely, no rate enhancement or increase in the stereoselectivity.
THE is more basic than diethyl ether and therefore the carbonyl group can
no longer compete for complexation of LiC104 .
When phenyllithium was allowed to react with 4-t-butylcyclo-
hexanone and 2-methylcyclohexanone (Tables 41 and 42, respectively)
in the presence of an equimolar amount of LiC10 4 in diethyl ether, only
a slight increase in the formation of the axial alcohols were observed
compared to the reaction without LiC104. When t-butyllithium was allowed
to react with the same two ketones in the presence of LiC104'
no change in
stereochemistry was observed since 100% axial alcohol was already formed
in the absence of LiC104' It was hoped that by increasing the steric re-
quirements in the transition state (eq. 5) some reversal of stereochem-
istry would ensue for those ketones which provided 100% of one alcohol
isomer. For example, the alkylation of norcamphor with CH3Li provides
100% of the endo-alcohol, but camphor provides 100% of the exo-alcohol
presumably due to the increased steric hindrance of camphor's methyl groups
encountered with exo attack. Unfortunately, the t-butyllithium reagent"
213
proved to be so bulky that the presence or absence of LiC10 4 is irrelavent.
An interesting observation concerning the reaction of t-butyl-
lithium with 4-t-butylcyclohexanone in the absence of LiC10 4 is that of
the products formed, 7% was the equatorial alcohol produced by reduction
of the ketone (Exp. 47) and 30% was recovered ketone. On the other
hand, in the presence of LiC10 4' the reduction product was only 1%, but
the amount of recovered ketone increased to 37%. From Table 43 it can be
seen that without the addition of LiC10 4 (Exp. 38), the reaction is
essentially over in 15 minutes, but in the presence of LiC104 (Exp. 39),
the reaction is over in less than 5 seconds. Evidently, the complexation
of the ketone with LiC104
increases the rate of the addition reaction
substantially without affecting the -hydrogen reduction pathway and
hence, 1,2-addition.product increases relative to reduction product in
the presence of LiC104 .
When t-butyllithium was allowed to react with 2-methylcyclo-
hexanone (Table 42, Exps. 34 and 35), 13% of the ketone was recovered
in the absence of LiC104
and 14% of the ketone was recovered when an
equal molar amount of LiC104 was present. When comparing this result with
that of the 4-t-butylcyclohexanone reaction, it is noticed that 30% of
4-t-butylcyclohexanone is recovered when LiC104
is not present. An
explanation of this result can be given in terms of enolization inter-
mediates. 4-t-Butylcyclohexanone has four hydrogen atoms available for
abstraction with the two equatorial ones being the most accessible.
2-Methylcyclohexanone has three hydrogens available for abstraction, but
19 the most stable enolate predominates in an 89:11 ratio (eq. 6). Based
on this, it is expected that the 4-t-butylcycloheicAnone would have
214
(6)
89%
roughly twice as many hydrogens available for abstraction than 2-methyl-
cyclohexanone, and this is indeed reflected in the amounts of recovered
starting ketones for the two reactions (30:13) after hydrolysis.
As discussed above, it was also observed that only the axial al-
cohol, cis-1,2-dimethylcyclohexanol, was produced whether LiC10 4 was pre-
sent or not. But, just as for the 4-t-butylcyclohexanone reaction (Table
43), the presence of LiC104 enabled the reaction to be complete in less
than 5 seconds for presumably the same reasons discussed above.
When phenyllithium was allowed to react with 4-t-butylcyclo-
hexanone, 58% of the axial alcohol was produced without the addition of
LiC104 . When LiC104 was added, the percentage of axial alcohol increased
to 69%. A similar result was obtained with the reaction of 2-methyl-
cyclohexanone and PhLi. The axial alcohol, cis-l-pheny1-2-methylcyclo-
hexanol, was produced in 88% yield without the addition of LiC10 4, but
with LiC104 added, the amount of axial alcohol increased to 94%. Evident-
ly, the reasons discussed for the CH3Li reactions also hold for the pheayl-
lithium reactions but not to the same degree. An explanation for this
maybe due to the fact that CH3Li exists as a tetramer and phenyllithium
as a dieter in diethyl ether. Therefore, the CB3Li could be considered a
215
more bulky reagent, and together with the effect of LiC10 4 has a greater
effect on the outcome of the addition reactions with ketones than phenyl-
lithium.
Organomagnesium Reactions
Chastrette and Amouroux20 reported that the presence of LiC104
or
n-Bu4 NC1 in the reaction of Grignard reagents with 4-t-butylcyclohexanone
has no effect on the stereochemistry of the products. Georgoulis et al. 21
also reported that the presence of KOBut has no effect on the stereo-
chemistry of the reaction of R2Ng compounds with 4-t-butylcyclohexanone;
however, it did increase the amount of enolization product. Our work in
this area was carried out in order to verify these results as well as in-
vestigate the influence of other salts and ketones in the reaction of
organomagnesium compounds with cyclohexanones. The results are recorded
in Table 44.
In general, we observed only slight increases (4-7%) in the forma-
tion of the axial alcohols for the reaction of dimethylmagnesium with
ketones in the presence of equal molar amounts of LiC10 4 compared to the
reactions in the absence of LiC104 ' However, for the norcamphor reactions
only the endo-alcohol was observed.
(aLue44, Exp. 48), was present in equal molar ratio with (CH3 ) 2Mg, the
stereochemistry of the product increased from 62% axial alcohol when no
salt was present (Exp. 42) to 75% (Exp. 48) when salt was present. A
similar result was obtained when MeMg0Pr i was allowed to react with the
same ketone. Therefore it maybe argued that a redistribution reaction
according to equation 7 took place which accounts for the so called salt
effect.
We also obserVed that when Mg(OPri)2
— Mg(OR) 2 Me2Mg V--- 2 MeMg(OR) (7 )
Other salts, NaBPh4 , LiBr or Lii, did not produce any different
results than the reactions carried out in the absence of salt. LiOBu t
did not increase the stereoselectivity of the reaction, but it did increase
the amount of enolization (30%) (Exp. 47) as was observed by Georgoulis 2l
using KOBut .
Of the alcohols produced from the reactions of 4-t-butylcyclo-
hexanone, 3,3,5-trimethylcyclohexanone and 2-methylcyclohexanone with
(CH3 ) 2Mg in the absence of LiC1O4, 62, 85 and 80% were the respective
axial or cis-alcohols. With LiC10 4 present, the amount of axial or cis-
alcohols increased to 69, 89 and 84%, respectively (Table 45, Exps. 50-57).
Granted, this is not much enhanced stereochemistry, but it is there.
However, the major effect of the addition of LiC104 to the reactions was
a pronounced rate enhancement. Therefore, a rate study was conducted with
and without LiC104
on the reaction of Me 2Mg with 4-t-butylcyclohexanone
at -78°C (Table 46). It was found that the presence of LiC104 increased
the rate 5 times over that observed for the reaction without LiC104'
It
can be reasoned as for the CH3 Li reactions that the increased polarization
resulting from the canplexation of the carbonyl group with LiC104 enabled
the rate of reaction to increase with a slight increase in enhanced stereo-
selectivity as well.,
Organoaluninum Reactions
Chastrette and Amouroux22
have reported that the addition of salts
such as NaF or KF or n-Bu4 NX where X = I, Br, Cl or F slowed down the
216
217
reaction of R3A1 compounds .with benzaldehyde in the order I >Br >C1 >F .
and NaF >1 >n-Bu4NF. In no case was the reaction faster in the.presence .
of a salt than in the absence of a .7,a1t!. They .I.ggested that this result
was probably due to the deactivating nature of a newly formed complex (C)
R
C=3
C
between R3A1 and the salt.2 3
As the complex becomes more stable the less
reactive it becomes.
With this data as precedence, we decided to study further any
change in stereochemistry of the products of reaction of Me 3A1 with
4-t-butylcyclohexanone in the presence of various salts in diethyl ether.
The results are summarized in Table 47. As in the other addition
reactions discussed above, only a slight increase (85% to 88%) in the
production of the axial alcohol was observed when the reaction was
conducted in the presence of LiC104 .
CHATTER IV
CONCLUSION
It was reported by earlier workers5
in this group that the addition
of CH3Li to 4-t-butylcyclohexanone in the presence of LiC10 4 -increased the
formation of the axial alcohol from 65 to 92% compared to the reaction
in the absence of LiC10 4' In the present study we conducted a more
detailed investigation of this unusual observation by allowing other
salts, ketones and organolithium compounds to react under these conditions.
It was observed that LiC10 4 had the greatest stereochemical effect on the
reaction of RLi compounds with cyclohexanones of all the salts tested.
When MeLi was allowed to react with 2-methylcyclohexanone in the presence
of LiC104'
only a modest increase in the stereoselectivity was observed.
But with 3,3,5-trimethylcyclohexanone and norcamphor under the same
reaction conditions produced no difference in the stereochemistry at all.
When phenyllithium was allowed to react with 4-t-butylcyclohexanone and
2-methylcyclohexanone a slight increase in the formation of the axial or
cis-alcohols was observed, but for the t-butyllithium reactions, no
difference was observed. The major result for all of these reactions was
a dramatic increase in the rate of reaction when the reaction are carried
out in the presence of LiC10 4 .
When Me2Mg was allowed to react with ketones in the presence of an
equimolar amount of LiC104 slight increases (usually 4%) in the amount
of axial alcohol were produced compared to the reaction without added
218
219
LiC104.
Again an increase in rate was observed. The increase in rate
for both the organolithitmt and Me2 Mg reactions was attributed to the
increased reactivity of the complex formed between-the carbonyl oxygen
and the lithium cation or in other words, the increased polarization of
the carbonyl group.
When Me3 AI was allowed to react with 4-t-butylcyclohexanone in the
presence of LiCIO4'
no change in the stereochemistry of the methylated
product was observed compared to the reaction without LiC10 4 . The addi-
tion of salts slowed down the reaction presumably due to deactivation of
Me3A1 as a result of complexation with the salt (Me3A1-LiC104).
220
Table 38.
EKP.
Reactions of .CH;1 -Metal Salts with 4-•lutylcyclohexanone in Diethyl Ether SolVent for 2 Hours at -78 °C in 2:1:1 Ratio.
AKLNI EQUATORIAL SALT ALCOHOL (%) a ALCOHOL (%) X'YIELD
b.
1 None 65 35 97
2 NaBPh4 68 32 97
3 Al(NPr2) 3 64 36 80
4 Al(OPri) 3 64 36 - 85
5 wocH3 ) 2 68 32 83
6 Mg(OPri) 2 69 31 85
7 Me4NI 70 30 92
8 LiBr 76 24 99
9 LiI 81 19 98
10 Li0But 75 25 73
11 LiC104
92 8 96
a) Normalized % Axial alcohol + % equatorial alcohol = 100%.
b) Yield was determined by glc and based on an internal standard. No ketone was recovered in any of the experiments.
14
15
16
17
18
19
0 65 35
1 92 8
0 100
1 100
88 12
1 98 4
100
1 100 0
12 I/
13
221
Table 39. Reactions of CH3Li-LiC104 with Various Ketones in Et 20 Solvent for 2 Hours at -78°C.
EXP. KETONE. LiC12.4 AXIAL EQUATORIAL KETONE ALCOHOL (%)a ALCOHOL (%) a
a) Normalized % axial alcohol % equatorial alcohol = 100%. No ketone was recovered in any of the experiments.
20 1 min 3 min 5 min
15 min No LiC10
4 30 min
60 min
22 I min 3 min 5 min
15 min No LiC10
4 30 min 60 min
Table 40. Rate of Reaction of Ketones with CH3Li-LiC10 4 at -78°C in Diethyl Ether Solvent.
RECOVERED AXIAL EQUATORIAL KETONE TIME' KETONE (%) ALCOHOL (%)a ALCOHOL'(%) a
222
EXP .
21 With LiC104 5 sec
10 sec
25 With LiC104 5 sec
10 sec
56 65 . 35 45 65 35 36 65 35 15 65 35 6 65 35 0 65 35
0 92 8 0 92 8
60 87 13 40 88 12 33 88 12 20 87 13 7 88 12 0 88 12
5 95 5 0 96 4
70 100 0 55 100 0 31 100 0 15 100 0 3 100 0 100
100 0 100
23 With LiC104
24 C
■1:::(9
No LiC104
5 sec 10 sec
1 min 5 min
15 min 30 min 60 min
120 min
a) Normalized % axial alcohol + % equatorial alcohol = 100%.
Table 41. Reactions of RLi-LiC10 /. Talp 4-t-Butylcyclohexanone in Et 20 Solvent for 2 Hours aC-78.C.
EXP. REAGENT LigP RECOVERED
KETONE (%)
ADDITION PRODUCTS REDUCTION AXIAL EQUATORIAL EQUATORIAL
ALCOHOL (%) 'ALCOHOL (%) ALCOHOL (%) KETONE
26 IleLi 0 0 65 35
27 1 0 92 8
28 t-BuLi 0 30 100 0 7
29 1 37 100 <1
30 PhLi 0 58 42
31 1 69 31
223
a) Normalized % axial alcohol + % equatorial alcohol = 100%.
Table 42. Reactions of RLi-LiC10 with 2-Methylcyclohekanone in Et 20 Solvent for 2 Haug at -78.C.
EXP. REAGENT LIC12 RECOVERED
KETONE (%) AXIAL
a
ALCOHOL (%) EQUATORIALa
ALCOHOL (%) KETONE
32 MaLi 0 88 12
33 1 0 96 4
34 t-BuLi 0 13 100 0
35 1 14 100 00
36 PhLi 88 12
37 1 0 94 6
a) Normalized % axial alcohol + % equatorial alcohol = 100%.
b) Without LiC10 4 1% of trans-2-,methylcyclohexanol was detected by glc.
With LiC104 only a trace of trans-2-methylcyclohexanol was detected by
glc.
224
EXP. KETONE
No LiC10 4
39 With L iC10
4
38
40
No LiC104
41 With LiC104
225
Table 43. Reactions of t-Butyllithium with Ketones in the Presence and Absence of LiC104 at -78°C in Et 2
0 Solvent in 2:1:1 Ratio.
TDIE RECOVERED KETONE. (%)
AXIAL EQUATORIAL ALCOHOL'(%) a—ALCOHOL'(%) a
5 min 45 100 0 10 min 35 100 0 15 min 30 100 0 30 min 30 100 0 60 min 30 100 0
120 min 30 100 0
5 sec 37 100 0 10 sec 37 100 0 60 sec 37 100 0 120 sec 37 100 0
2 min 50 100 0 5 min 35 100 0
10 min 26 100 0 15 min 13 100 0 30 min 14 100 0 60 min 13 100 0
120 min 14 100 0
5 sec 13 100 0 10 sec 13 100 0 60 sec 13 100 0
120 sec 13 100 0
a) Normalized % axial alcohol + % equatorial alcohol = 100%.
Table 44. Reactions of Me 2Mg-Salt with 4-t-Butylcyclohexanone in Et20 Solvent for 2 Hours at -78°C in 2:1:1 Ratio.
EKP. AKIAL
SALT • 'ALCOHOL'(%) 4 EQUATORIAL
ALCOHOL . (%)a 1- YIELDb
42 None 62 38 96
43 LiC104 65 35 96
44 NaBPh4 60 40 97
45 LiBr 65 35 95
46 LiI 65 35 94
47 Li0But 65 35 70
48 Mg(OPr i) 2 74 26 89
49 Metig0Pr i (only) 75 25 91
a) Normalized % axial alcohol + % equatorial = 100%.
b) Yield was determined by glc using an internal standard. No ketone was recovered in any experiment.
226
50
51
52
53
54
55
56
57
Ling 'KETONE
RECOVERED AXIAL KETONE (70"AICOHOL'(%) a
EQUATORIAL ALCOHOL'(%) 4
0 0 62 38
0 62 31
0 0 85 15
0 89 11
0 0 80 20
1 0 84 16
0 0 100
100 0
'EXP. " KETONE '
227
Table 45. Reactions of o Me
2 Mg -LiC10, with Ketones in Et 20 Solvent for 2 •
Hours at -78 C in 2:1%1 gatio.
a) Normalized % axial alcohol + % equatorial alcohol = 100%.
228
Table 46. Reactions of Me „)Mg with 4-t-Butylcyclohexanone in the Presence and Absence of tiC10 4 in Et2 0 Solvent at -78 °C in 2:1:1 Ratio.
EXP. CONDITIONS RECOVERED KETONE (%)
AXIAL ALCOHOL (%) a
EWATORIAL ALCOHOL (%) a
58
59
No LiC104
LiC104 Added
2 min
5 min
10 min
15 min
30 min
60 min
120 min
10 sec
20 sec
40 sec
60 sec
120 sec
120 min
50
35
20
13
55
0
0
0
0
0
0
0
62
62
61
62
63
62
62
69
70
70
68
69
69
38
38
39
38
37
38
38
31
30
30
32
31
31
a) Normalized % axial alcohol + % equatorial alcohol = 100%.
Table 47. Reactions of Me,4A1-Salt with 4-t-Butylcyclohexanone in Et20
Solvent for 12 Pours at -78 °C in a 2:1:1 Ratio.
RECOVERED AXIAL EQUATORIAL • EXP. SALT KETONE (%) ALCOHOL (%) a ALCOHOL .(%) '7,'YIELD w
229
60 None
85 15 98
61 LiC104 88 12 96
62 NaBPh4 0
88 12 95
63 LiBr
0
85 15 95
64 LiI
0
85 15 94
65 Li0But
85 15 72
a) Normalized % axial alcohol + % equatorial alcohol = 100%.
b) Yield was determined by glc using an internal standard.
LITERATURE CITED
1. T. L. MacDonald and W. C. Still, J. Am. Chen. Soc., 97, 5280 (1975).
2. D. P. Novak and T. L. Brown, J. Am. Chem. Soc., 94, 3793(1972).
3. P. West and R. Waack, J. Am. Chem. Soc., 89, 4395(1967).
4. R. G. Pearson and C. D. Gregory, J. Am. Chem. Soc., 98, 4098(1976).
5. E. C. Ashby, J. J. Lin and J. J. Watkins, J. Org. Chem., 42, 1099 (1977).
6. E. C. Ashby and J. J. Watkins, J. C. S. Chen. Comm., 784(1976).
7. E. C. Ashby and J. J. Lin, Tetrahedron Lett., 1 709( 1 977).
8. A. Pull and D. Poolock, Trans. Faraday Soc., 5115 11(1958).
9. D. F. Schriver, "The Manipulation of Air-Sensitive Compound", McGraw-Hill, New York, New York, 1969.
10. E. C. Ashby and R. D. Schwartz, J. Chem. Educ., 51, 65(1974).
11. H. Gilman and A. H. Haubein, J. Aa.Chem. Soc., 66, 1515(1944).
12. E. C. Ashby and R. C. Arnott, J. Organomet. Chem., 14, 1(1968).
13. E. C. Ashby, J. R. Sanders, P. Claudy and R. Schwartz, J. Am. Chem. Soc., 95, 6485(1973).
14. E. C. Ashby, J. Laemmle and P. V. Roling, J. Org. Chem., 38 2526 (1973).
15. a) O. L. Chapman and R. W. King. J. Am. Chem. Soc., 86, 1256(1964); b) R. J. Ouellette, J. Am. Chen. Soc., 86, 4378(1964); c) G. D. Meakins, R. K. Percy, E. E. Richards and R. N. Young, J. Chem. Soc. C, 1106(1968); d) J. Battioni, W. Chodkiewicz and P. Eadiot, C. R. Icad. Sci., Ser. C, 264 991(1967); e) J. Battioni, M. Chapman and W. Chodkiewicz, Bull. Soc. Chim. Fr., 976(1969); f) J. Battioni and W. Chodkiewicz, Bull. Soc. Chim. Fr., 981(1969); g) 3. Battioni and W. Chodkiewicz, Bull. Soc. Chim. Fr., 1824(1971).
16. J. R. Luderer, J. E. Woodall and J. L. Pyle, J. Org. Chem., 36, 2909(1971).
230
231
17. J. Ficini and A. Maujean, Bull. Soc. Chin. Fr., 219(1971).
18. This work was orginall .y done by J. J. Lin and checked by the author.
19. H. O. House, W. L. Roelof s' and B. M. Trost, -J. Org. Chem., 31, 646 (1966).
20. M. Chastrette and R. Amouroux, Bull. Soc. Chin Fr., 1955(1974).
21. C. Georgoulis, B. Gross and J. C. Ziegler, C. R. Acad. Sci., Ser. C, 273, 378(1971).
22. M. Chastrette and R. Amouroux, J. Organomet. Chem., 40, C56(1972).
23. M. Chastrette and R. Amouroux, J. Organomet. Chen., 70, 323(1974).
233
CHAPTER . I
INTRODUCTION
Background
It is well known that LiA1H4 favors 1,2-reduction of enones.
1
the other hand, the reactivity of LiA1H 4 can be substantially modified by
the addition of metal salts. In this connection LiA1H4-A1C13 has found
unusual applicability in epoxide reductions,2 LiA1(OCH3 ) 3E-CuI can effect
reductive removal of halo and mesyloxy groups 3 and LiA1H4-TiC13 has been
found to be an excellent coupling reagent.4 The LiA1H
4-CuI5 reagent has
been found to reduce enones conjugatively in 98% yield with 100% regio-
selectivity. However, it was found that the reactive intermediate was
H2A1I and not CuH or CuA1H4.
Recently there has also been an increased interest in methods for
effecting 1,4-conjugate addition to a,-unsaturated systems. 7 In addi-
tion to lithium dialkylcuprate and copper-catalyzed Grignard reagents,
Brown and Kabalka8 have found that trialkylboranes undergo 1,4-addition
to a variety of (10B-unsaturated substrates via a free radical chain pro-
cess. More recently Kabalka and Daley9 found that trialkylaluminum com-
pounds exhibit analogous behavior when photolyzed at -78°C, or in the
presence of catalytic amounts of oxygen, and were able to demonstrate the
intermediacy of free radical species. Ashby and Heinsohn10
and Mole, et
al. 11 independently have shown that nickel acetylacetonate does catalyze
the 1,4-addition of selected enones in high yields and regioselectivity.
Taking into account that the active species in the 1,4-reduction
of enones by LiA1H4-CuI is H2A1I, it seems quite within reason to inves-
tigate the possibility of performing the 1,4-addition of enones with sub-
stituted alkylaluminum compounds without catalysts.
Purpose
As stated above, earlier workers in this group have shown that
H2A1I provided 100% of the 1,4-conjugate addition product when allowed to
react with enones. Therefore we investigated the possibility of using
R2A1X and RA]X2 compounds to promote the non-catalyzed 1,4-conjugate
addition to enones. We also wanted to conduct a systematic study con-
cerning these compounds (e.g.. solvents, molar ratios and temperature
effects) towards the alkylation of model ketone systems. To achieve
these goals we prepared a varied array of substituted alkvlaluminum
compounds and allowed them to react with representative enones and ketones
under consistent conditions.
234
235
CHAPTER II
EXPERIMENTAL SECTION
General Considerations
Manipulations of air-sensitive compounds were performed under
nitrogen in a glove box equipped with a recirculating system using man-
ganous oxide columns to remove oxygen and dry ice-acetone traps to remove
solvent vapors. 12 Reactions were performed under argon or nitrogen at the
bench using Schlenk tube techniques. 13 Syringes equipped with stainless
steel needles were used for transfer of reagents. All equipment was
flash flamed or heated in an oven and cooled under a flow of nitrogen or
argon before use. Proton NMR spectra were obtained at 60 MHz using a
Varian A-60 or T-60 NMR spectrometer. GLPC analyses were obtained with a
Hitachi (Perkin-Elmer) Model RMU--7 or Varian Model M-66 mass spectrometer.
The it spectra were determined with a Perkin-Elmer, Model 621 or Model 257
infrared recording spectrophotometer.
Analytical
Active CH3 or C2H5 group analysis were carried out by hydrolyzing
samples with hydrochloric acid on a standard vacuum line equipped with a
Toepler pump.16
Aluminum was determined by adding excess standard EDTA
solution to hydrolyzed samples and then back titrating with standard
zinc acetate solution at pH 4 using dithizone as an indicator. Halide was
determined by titration with AgNO 3 and back titration by KCNS with ferric
alum indicator.
236
Materials
Fisher Reagent Grades anhydrous diethyl ether and tetrahydrofuran
(THF) were distilled from LiA1H4 and NaA1H4'
respectively prior to use.
Lithium aluminum hydride solutions were prepared by refluxing
LiA1H4 (Alfa Inorganics) in THF or diethyl ether for at least 20 hours
followed by filtration through a glass fritted funnel aided by Celite
filter aid in the dry box. The clear solution was standardized for
aluminum content by EDTA titration.
2,2,6,6-Tetramethyl-trans-4-heptene-3-one, m. p. 43.0-43.7 °C, NMR:
(CC14'
TMS), 6.2-7.0 ppm (2 H, quartet, olefinic); 1.10 ppm (18 H, singlet,
two t-butyl groups) was obtained from co-workers, J. R. Boone, T. L.
Wiesemann and J. J. Lin. Trans-3-penten-2-one, chalcone (Aldrich), 4-t-
butylcyclohexanone (Frinton), 2-methylcyclohexanone and 3,3,5-trinethyl-
cyclohexanone (Chemical Samples) were obtained commercially and purified
by sublimation or distillation under vacuum.
Diisopropyl amine (Fisher) was purified by distillation over NaOH.
Tert-butyl alcohol (Fisher) was purified by distillation over CaH2 .
2,2,6,6-Tetramethylcyclohexanol was prepared by the LiA1H 4
reduction at 0°C in diethyl ether of 2,2,6,6-tetramethylcyclohexanone
which was prepared by the exhaustive methylation of cyclohexanone in the
presence of NaH-t-Bu0H-MeI 17 (See. Part III).
Trimethylaluminum and triethylaluminum were obtained from Texas
Alkyls and distilled under vacuum in a dry box. Triphenylaluminum was
prepared according to the method of Mble.14
Aluminum foil (10 g) activated
by (aq) HgC1 2 and washed with ethanol and diethyl ether and was stirred
under vigorous reflux with diphenyInercury (Alfa Inorganics) (30 g) in
237
benzene for 4.5 hours. The solution was filtered and the solid residue
was found after treatment with dilute acid to contain mercury. The
filtrate was evaporated to a volume of 100 ml. The benzene solution of
triphenylaluminum was analyzed for aluminum, content by EDTA titration and
back titrated with zinc acetate with dithizone as the indicator.
Me2A1C1, MeA1C12 , Me2A1Br, MeAlBr 2 , Me2A1I, MeA1I2 , Et 2A1C1, EtA1C12 ,
Et2A1Br, EtA1Br 2 , Et2A1I, EtA1I 2 , Ph2A1I and PhAlI2 were prepared by the
distribution reactions of Me3 Al, Et3 Al or Ph3Al with freshly sublimed and
finely crushed A1C13 , A1Br3 or A113 (Alfa Inorganics) in THE or benzene at
0-10°C with good stirring. 15 The iodo-compounds were also prepared by
adding a stoichianetric amount of iodine in benzene or THE to Me 3A1 ' Et3A1
or Ph3A1 at 0°C. The resulting methyl or ethyl iodide (eq. 1) was removed
I2 R3A1 bent
R2A1Ior THE (Ph) 3 (Ph) 2
RI (1)
(PhI)
by applying a partial vacuum. If any solvent was lost, it was replenished
with freshly distilled solvent. The iodobenzene did not appear to inter-
fere with the subsequent reactions since the phenylaluminum compounds
prepared both ways provided the same result.
Me2A10But , Et2A10But , Ph2A10But , Me2A1OCH3 , Et 2A1OCH3 , Ph2A10CB3 ,
Me2A101‘ , MeA1( 2' 2 Et A10. , 7) Ph2 , Me2Al(NPr i)
2
Et2A1(NPr2) and Ph2A1(NPr 2 ) were prepared by the addition of the appropri-
ate alcohol or amine in THE or benzene and allowed to react in stoichio-
metric amounts with the appropriate trialkyl- or arylaluminum. compound.
All the reagents were analyzed for aluminum content by the usual EDTA
238
titration. The methyl and ethyl derivatives 'were also analyzed by normal
gas techniques using a standard vacuum line and a Toepler pump. 16
General Reactions of Enones
A 10 X 18 mm test tube equipped with a Teflon coated magnetic
stirring bar was dried in a oven or flash flamed and allowed to cool under
nitrogen or argon flush and sealed with a rubber serum cap which was
connected by means of a syringe needle to a nitrogen-filled manifold
and a mineral oil filled bubbler. The alkyl or aryl aluminum reagent
(ca. 0.1-0.5 mmole) was syringed into the test tube, and then the calcu-
lated amount of enone (in THE or benzene solvent with an internal standard,
n-C12H26 or 27C 14H30) was added to the stirred reagent at the desired
temperature. After the designated reaction time, the reaction was
quenched slowly with H2O and dried over MgSO 4 .
Methylation of Enones
The methylation products were determined according to the procedure
of Ashby et al 18 A 10 ft. 5% Carbowax 20K on Chromosorb W-NAW column
at a flow rate of 55 ml of He/min was used to separate the 1,4- and 1,2-
methylation products of 2,2,6,6-tetramethyl-trans-4-hepten-3-one (enone I)
(120°C), 3-penten-2-one (enone II )(90°C), chalcone (enone III) (210°C)
and 2-cyclohexenone (enone IV) (125 °C). Authentic samples of the 1,4-
and 1,2-methylation products were obtained from co-workers J. J. Lin and
T. L. Wiesemann. The 1,2-methylation products were prepared from
the reaction of the enone with CH 3Li, and the 1,4-methylation products
were prepared from the reaction of the eneone with LiCu(CH 3 ) 2 . The per
cent yield for each reaction was normalized to 100% = enone recovery % +
1,2-product % + 1,4-product %. Retention times of products varied
239
slightly depending on glc conditions for enones I and II, but the order of
elution was always the same: enone first, 1,4-methylation product second
and 1,2-methylation product last. However, when 2-cyclohexenone was the
substrate, n-octyl alcohol was employed as the internal standard and the .
order or retention was 1,2-methylation product first, 1,4-methylation
product second and the enone last. The products from the chalcone reaction
were determined by 1H NMR: enone (2 H, vinyl, multiplet at 6.7-7.4 ppm);
1,2-methylation (3 H, methyl group, singlet at 1.43 ppm and 2 H, vinyl,
multiplet at 6.5-7.5 ppm); 1,4-aethylation (2 H, methylene group, doublet
at 2.7 ppm J = 6 H z , 3 H, methyl group, doublet at 1.15 ppm J = 6 H z and
1 H, methine, multiplet at 2.80 ppm.
Ethylation of Enones
The ethylation products from the reaction of the RnAlK3-11 compounds
with 3-penten-2-one (enone II) were deterained on a 10 ft. 5% Carbowax 20M
on Chromosorb W-NAW column at 130 °C with a flow rate of 45 ml of He/min.
The order of elution was 1 2-reduction product, enone, 1,4-ethylation pro-
duct and 1,2-ethylation product. An authentic sample of the 1,2 reduction
product, 3-penten-2-ol, was obtained from co-worker J. J. Lin. Pfaltz and
Bauer Chemical Company provided an authentic sample of the 1,4-ethylation
product, 4-methyl-2-hexanone. The 1,2-ethylation product, 3-methyl-4-
hexen-3-ol, was prepared from the reaction of triethylaluminium or ethyl-
magnesium bromide with the enone. 19
Phenylation Of Enenes
The phenylation products from the reaction ofPhnAEC3-n compounds
with 3-penten-2-one (enone II) ware determined on a 10 ft. 5% Carbowax 20M
240
On Chromosorb W-NAW column at 1600C with a. flow rate of 45 ml of He/min.
The order of elution was enone first, L,4-phenylation product and then the
1,2-phenylation product. The 1,2-phenylation product, 2-phenyl-3-penten-2-
ol, was obtained from the reaction of phenyllithium with 3-penten-2-one and
from the Chemical Samples Company. The 1,4-phenylation product, 4-phenyl-
2-pentanone, was isolated by prep glc and analyzed according to the pro-
cedure of Melpolder and Heck.26
General Reactions with Ketones
The same general procedure used for the enone reactions described
above was followed for the ketone reactions.
Methylation of Ketones
The MenAlK3-n
compounds listed above were allowed to react with
4-t-butylcyclohexanone (ketone I), 3,3,5-trimethylcyclohexanone (ketone II)
or 2-methylcyclohexanone (ketone III) in benzene or THE at various teu-
peratures. After the designated reaction time, the reaction with an
internal standard, n-C 14H30, was quenched slowly with H 2O and dried over
M.004 . A 12 ft. 10% FFAP on Diatoport S column (column temperature: 150°C,
helium flow rate: 60 ml/min) was used to separate the products for the
4-t-butylcyclohexanone reaction. The retention time was 13.4 min. for n-
C14H30 ' 32.7 min. for cis-1-methy1-4-t-butylcyclohmcAnol 38.0 min. for
4-t-butylcyclohexanone and 42.0 min. for trans-1-methy1-4-t-butylcyclo- __
hexanol. A 12 ft. 10% diglycerol on Diatoport S column at 80 °C was used
to separate the products from the 2-methylcyclohexanone reaction. The
retention time was 4.4 min for the ketone, 5.2 min. for cis-I-methyl-2-
methylcyclohexanol, 9.5 min. for trans-l-methyl-2-methylcyclohexanol and
16.0 min. for n-C 14 H30' A 10 ft. 20% SAID on Chromosorb W column at 180°C
(flow rate was 60 ml of He/min) was used to determine the products from the
241
methylation reactions of 305-trimethylcycloheXanOne (ketone III). These
conditions gave retention times of 5.0, 4.0 and 6.0 minutes for the ketone,
axial alcohol and equatorial alcohol, respectively. Authentic samples of
all products were obtained from co-worker J. J. Lin.
Ethylation of Ketones
The ethylation products obtained from the reaction of Et n i-n
compounds with 4-t-butylcyclohexanone (ketone I) were determined by a 10
ft. 20% SAIB on Chromosorb W column at 150 °C (flow rate of 60 ml of
He/min). The retention times for ketone (I), axial alcohol (alkylation),
equatorial alcohol (alkylation), axial alcohol (reduction) and equatorial
alcohol (reduction) were 30.0, 39.0, 45.0, 28.0 and 32.0 minutes, respec-
tively.
Phenylation of Ketones
The products and their ratios from the reactions of Ph A1K with n 3-n
4-t-butylcyclohexanone (ketone I) were determined by the procedure of
Ashby, Laemmle and Roling.21
After the desired reaction time, the
reactions conducted in benzene were subjected to vacuum until all benzene
had been removed. Wet diethyl ether was then added to the carbinolate in
order to effect hydrolysis. The solution was then transferred to a
separatory funnel and the aluminum salts were removed by several washings
with distilled water. The ether layer was separated and allowed to evap-
orate and DMSO-d6 was added to the sample. The sample was then dried
over Linde 4A molecular sieves and transferred to a M1R tube. In those
cases where THE or diethyl ether was employed as a solvent, the work-up
was identical except that the solution was hydrolyzed directly with dis-
tilled water. In the case of the phenylation of 4-t-butylcyclohexanone
242
(ketone I), the chemical shifts are 4.56 ppm and 4.73 ppm for the axial
and equatorial hydroxyl protons, respectively. The assignments of each
alcohol hydroxyl peak to a particular isomer were based on numerous reports
in the literature concerning their chemical shifts in DMSO and DMSO-d 6.22
CHAPTER III
RESULTS AND DISCUSSION
Enone Reactions
methylatiot o f'EnOries
Four enones, 2,2,6,6-tetramethylttans-4-hepten-3-one (I), 3-
penten-2-one (II), chalcone (III) and 2-cyclohexenone (IV), were chosen
to react with Me3A1, Me2A1C1, Me2A1Br, Me 2A1I, MeklI2 , Me2A1OCH3 , Me2A10But
243:
i and Me2A1NPr 2 in THF and benzene.
It was expected that for steric reasons the MeA1X2 compounds
would be more regioselective than Me 2A1X or Me3A1 and the regioselectivity
of the addition would decrease as the steric requirement of the X group
decreases (I > Br > Cl or 0 > OBut > OMe).
Earlier workers in this group chose 2,2,6,6-tetramethyl-trans-4-
heptene-3-one (enone I) as the representative enone for the regioselective
reduction study with linA1X3_n compounds. Therefore enone (I) was also
chosen as the representative enone for this study. The reagents mentioned
above were freshly prepared in THF or benzene (Experimental Section) for
each reaction and allowed to react with enone (I). The results are shown
in Table 48. Very little difference was observed in the results for the
reactions in THF or benzene except that in most cases, less starting
material was recovered when the reactions were conducted in benzene than
in THF. Enone (I) was allowed to react with Me3A1 which resulted in no
1,4-addition at methide:enone = 3. However at bethide:enone = 1, small
amounts of 1,4-addition (1-5%) were observed. This was expected since the
steric requirement of the reagent increases from Me 3A1 to MezAlOR to
MeAl(OR) 2 during the course of the reaction. It is clear from Table 48
that as Cl, Br or I replace a methyl group in the reagent, the amount of
recovered enone (I) increases indicating rate retardation because of the
increasing=steric bulk of the reagents. However, as the rate decreases
the amount of 1,4-addition increases in the following order: Me3A1 (1-5%)<
.MeA1C1 (5-7%) < Me 2A1Br (20-21%) < Me2AlI (98-99%) > MeA1I 2 (95%).
Therefore we can say that the greater the steric bulk, the slower the
reaction but the greater the stereoselectivity of the reagent.
However, due to the steric requirement of MeA1I2 , the addition with enone
(I) is much slower compared to Me 2AlI. Because MeAlI 2 reacts so slowly,
the regioselectivity suffers slightly probably due to a small amount of
A1tle3 formed by disproportionation (eq. 1).
3 Met1'2
2 AlI3 Me3A1 (1)
Earlier workers in this group have shown that 1,4 conjugate reduo-
tion of enones with H2A1X compounds resulted from a six-center transition.
, state (A). 5 This type of transition state would make it more difficult
244
A e
H2O (2)
245
for the rigid cyclohexendne systems, cis-enoneg and trans-enones possessing
disubstitution at the 8-carbon of the enone to accomodate such a transition
state (A) and hence theSe kinds of compounds should react more slowly.
Presumably, the 1,4-conjugate addition which took place in the above re-
actions also proceeded through an analogous transition state and this
would explain the rate retardation encountered with bulky reagents or
reactants.
When the alkoxy or dialkylaniino reagents were allowed to react with
enone (I) only starting material was observed with a decrease in mass ba
lance (70-91%). It was expected 'that Me2A10 MeA1(0
Me2AlNPr
2 and MeAl(NPr
2) 2 which are very bulky reagents would be very
selective towards 1,4-addition. However, only starting material (enone I)
was observed after being allowed to react with these reagents accompanied
by a decrease in mass balance (Table 48).
As Ashby and Heinsohn10 noted, 1.4-addition can take place forming
an intermediate enolate which can then react with starting enone (eq. 2)
+ Me3A1
Cat
246
when trimethylaluminum is allowed to react with -,-cyclohexenone in the
presence of a catalyst (nickel acetylacetonate). A similiar sequence
is proposed for the Mbn AIX3-n
reactions which can account for the loss of
mass balance.
Since Me2A1I produced the greatest amount of 1,4-addition product
when allowed to react with enone (I), it was decided to allow Me 2AlI to
react with other enones i .e.g. 3-pentene-2-one (II), chalcone (III) and
2-cyclohexenone (IV) (Table 49). With enone (II), a 55:45 ratio of 1,4:
1,2 addition products in THF and a 61:39 ratio in benzene were obtained.
Similar results were observed for enones (III) (67:33) and (IV) (65:35);
however, for enone (IV) a larger amount of enone was not accounted for
presumably due to further reaction of the intermediate enolate as dis-
cussed above.
Since the bulky reagents, e..&. MeA1( ) 2' were not effective
for 1,4 conjugate addition to enones, we decided to add bulky coordinating
agents, e.g. Ph3P, n-Bu3P or HMPA to Me3A1 and then allow the mixture to
react with enones (I) and (II) (Table 50). It was found that enone (I)
reacted very slowly in THF for all of the reactions. In benzene, however
Ph3P allowed only 64% of enone (I) to produce the methylated products with
8% being of the 1,4 orientation, while n-Bu3P allowed only 6% of the enone
to react producing a 15:85 ratio of 1,4:1,2 products. No methylated
products were observed when HMPA was present.
For enone (II), the addition of coordinating agents listed above
increased enolization such that in no case was more than 4% of enone (II)
recovered. But with HMPA in THF and benzene, the 1,2-1methylation product
was obtained in 32 and 47% yields, respectively.
247
These results do not offer much encouragement for the development
of stereoselective reagents towards the methylation of enanes.
Ethylation of Enones
To continue our study of non-catalyzed 1,4-addition of organo-
aluminum compounds, we allowed a more bulky alkyl group, ethyl, to be
present on the aluminum atom. Hence, we allowed Et n AM
i-n compounds to
react with an unencumbered enone, 3-penten-2-one (enone II), in THE or
benzene at different tines, temperatures and mole ratios of reagent to
enone. The results of this study are listed in Table 51.
When Et3Al was allowed to react with enone (II), the 1,2-ethylation
product, 3-methy1-4-hexen-3-ol, was the major product with a small amount
("2%) of the 1,2-reduction product, 3-penten-2-ol, also being produced.
As one of the ethyl groups was replaced with Cl, Br, or I, the amount of
1,2-reduction product also increased from approximately 3 to 6 to 13%,
respectively, in both TRF and benzene. However, it should be noted that
the benzene reactions always produced less ethylation and more reduction
product than the THE reactions. Also, the amount of 1,4-ethylation pro-
duct also increased from approximately 1 to 3 to 18% as one of the ethyl
groups was replaced with Cl, Br or I, respectively. If two iodo groups
were introduced into the system, the amount of 1,4-ethylation increased to
r28%, but the 1,2-reduction product also increased to 35%. These results
are reminiscent of the MenAIX3-n reactions discussed above.
An observation for this reaction which differed from the MeriAlX3_ ,11
reactions was that in no case was starting enone recovered. However, the
mass balances were approximately the same which implies that the alkylation
reactions take place at about the same rate for both systems, but the 8-
248
reduction pathway in the EtAIX system. is prominent because of the in-
creased steric requirements.
If more bulky groups (e.I.NPr 2 , 0But , or
duced into the system, the amount of 1,4-addition or 1,2-reduction was
not increased, but the-mass balance was lowered considerably (32-42%)
Which was also observed for the MenAIX3-n
reactions. This result may
also be attributed to the already discussed enolization reactions.
Phenylation of Enones
When triphenylaluminum, a bulky reagent, was allowed to react with
an unencumbered enone such as 3 -penten-2-one (II) (Table 52), 60% of the
1,4-phenylation product was observed via glc. As other Ph 1AIX3n com-
pounds were allowed to. react with enone (II), an increase in the production
of 1,4-addition product was observed for X = C1(65%), Br(90%) and I(100%)
but accompanied by a lower mass balance (approximately 77, 75 and 66% in
both THE and benzene). This May beexplained on the basis that as the
amount of 1,4-addition increases, the newly formed enolate can attack
starting enone (eq. 2). This explanation accounts for the observation of
no recovered enone and is confirmed by the fact that when PhA1I 2 was allowed
to react with enone(II) 100% of the 1,4-addition product was produced but
only in 'N, 10% yield. The two large iodine atoms slow down the reaction
considerably, and, therefore, the newly formed enolate has plenty of time
to react with starting enone and thereby lower the yield of 1,4-addition
product.
If other balky substituents NPr2' 0Bu
t, and were in-
) were intro-
troduced into the system, 74, 76 and 81% of the 1,4-addition product was
249
observed, respectively. However, just as for the other reactions
described above, only a small yield (approximately 25, 22 and 11%, respec-
tively) of the phenylated product was obtained with no starting enone
recovered. Direct enolization or the reaction of the newly formed 1,4-
enolate with starting enone may be responsible for this low mass balance.
Ketone Reactions
Methylation of Ketones
A most unusual observation was made when it was found that Me 3 Al
attack is favored by 85%. 24,25 It was shown that in hydrocarbon sol-
vent "compression effects" controlled the product distribution when
reagent to ketone ratios were 2:1 of greater.21
In ratios less than 2:1
or in THF, steric effects were the effective force. Therefore, Me nAlK3-11
compounds were allowed to react with ketones with bulky substituents in
order to observe their effectiveness towards stereoselective addition.
The following reagents were allowed to react with a
compounds were allowed to react with a representative ketone, 4-t-butyl-
cyclohexanone, in order to observe their effectiveness towards sterero-
selective addition.
The following reagents were investigated in THF and benzene at
different times, temperatures and mole ratios of reagent to ketone:
Me3A1, Me AlC1, Me 2A1Br, Me2A1I, MeA1I2 , Me 2A1NFr2 , Me2A1OCH3 , Me2A10But ,
Me2Al , and MAA1( ) 2 . The results of this study are listed in
Table 53. It was found that increasing the steric bulk of the methyl-
aluminum halide reagents from Cl. to Br to I had little effect on the ster-
eoselectivity of the reaction. However, the increased steric bulk of the
250
reagent decreased the rate of reaction and therefore, 80, 83 and 97% of
the ketone was recovered according to glc analysis.
However in benzene, a reagent/ketone ratio of 3:1 also had little
effect on the stereoselectivity (6-10% axial alcohol), but as the halo
substituents changed from el to Br to I the yield of addition product
decreased from 95% for Me3A1 to 79% to 66% to 53%, respectively which
reflects the increased steric bulk of the reagents. Also, the amount of
recovered ketone increased from 0% to 17% to 25% to 40% respectively.
This also reflects the increased steric bulk by decreasing the rate of
reaction.
At .a reagent/ketone ratio of 1:1 in benzene the increase in the
bulkiness of the halo substitaent decreased the rate of reaction as was
indicated by the amount of recovered ketone (e.g. Me 3A1 (33%), Me2A1C1
(70%), Me2A1Br (80%) and Me 2A1I (87%)). Also, the amount of axial alcohol
decreased from Me3A1 (75%), to Me 2A1C1 (57%) to Me2A1Er (40%) to Me2A1I
(20%). Therefore, it may be assumed, that the increased bulkiness of the
reagent does indeed increase the stereoselectivity of the reagent.
If two iodo groups are introduced into the reagent, MeAlI 2' the rate
is slowed down even more in both THF and benzene resulting in 97% recovery
of ketone (I) in THF, 91% recovery in benzene when the reagent/ketone ratio
was 1:1 and 50% when the ratio was 3:1, results which compare to 40% ob-
tained for Me2AlI under the same conditions. The production of axial al-
cohol remained around 6% in a 3:1 ratio of reagent to ketone in benzene,
but in a 1:1 ratio the axial alcohol decreased to 18% which was only a 2%
decrease from that observed for the Me2A1I reaction. Therefore, it may be
safe to assume that Me2AlI is the most stereoselective reagent studied,
251
but with only 6% of the ketone inethylated it is not a practical reagent.
- , • , e
If other bulky groups (ca. NPr 2 OCH3 , OBu , or ( ) 2)
were introduced into the reagent, the major product obtained from the
reaction of these reagents with ketone (I) was recovered ketone. There-
fore it maybe assumed that the reaction was very slow or enolization took
place as was indicated by a very low mass balance for the MeA1( ) 2
reaction (0-8%). The amount of addition product ranged from 0% to 20%.
In all cases the major alcohol product was the axial alcohol (80-100%).
The above reagents represented attempts to influence the distri-
bution of products by steric bulk of the reagents. It was also of impor-
tance to observe the effect when strong bulky coordinating solvent -4, e..8.L.
Ph3F, n-Bu3P, HMPA or DME, were introduced with the ketone and then al-
lowed to react with trimethylaluminum (Table 54). No significant change
in stereochemistry from that of Me3A1 alone was observed, however enoli-
zation increased and the reaction slowed down considerably as indicated
by a large amount of recovered ketone.
Ethylation of Ketones
To continue our study of ketone addition reactions, we allowed a
representative ketone, 4-t-butylcyclohexanone, (Table 55) to react with
Et A ]X3-n compounds where X = Cl, Br or I. The amino- and alkoxy-reagents n
were not prepared since enolization was the main reaction for the MenAlK
3-n
compounds. Also , since the stereochemistry of the alcohol products from
the same reactions in THE remained almost the same thoughout the study,
the Et II Aix.. compounds were decided to be carried out in benzene only. J-n
For the reaction of Et3A1 with ketone (I) in a 1:1 ratio, 46% of
ketone (I) was recovered, but 50% of the ketone was either ethylated (64%)
252
or reduced (36%). Of the ethylated products, 79% was the axial alcohol
while 81% of the reduced product was the equatorial alcohol. As one Cl,
or Br or I group was introduced: in the reagent under these conditions,
the amount of , recovered ketone remained essentially the same (`L30%).
This was a decrease of roughly 16% from the Et3A1 reaction. Also, the
amount of ethylated product decreased to 41, 40 and 43%, respectively,
compared to 64% from the EtaAl reaction. Therefore, ona can conclude
that with the introduction of halogens, the t3-reduction pathway becomes
prominent. Also, for the ethylated product, there was an initial decrease
in the percent of axial alcohol produced from the Et3A1 reaction compared
to Et 2'&ICI. But, then there was a steady increase in the axial alcohol
from Cl to Br to I (65 to 70 to 75%, respectively). However, the per
cent of the reduced axial alcohol remained constant at 20%. This reflects
the increasing steric bulk of the reagents.
For the 3:1 ratio of reagent to ketone reactions, I), 6% of the
ketone was recovered for the Et 2A1X reactions compared to no recovered
ketone for Et3Al. The amount of ethylated and reduced products (40-43%
and 57-60% respectively) remained constant throughout the halide series.
but the amount of axial ethylated alcohol decreased from C1(27%) to
Br(20%) to 1(15%), again reflecting the increasing steric bulk of the
reagent. The reduced axial alcohol remained constant for the chloride
(26%) and bromide (25%) reactions but increased for the iodide reaction
(35%).
The trends observed for the mono halide reactions were amplified
for the EtA1I2 reactions because of the greatly increased steric bulk of
the reagent. That is, a-larger amount of ketone was recovered for both
253
the 1:1 ratio and 3:1 ratio xof reagent Lo ketore reactions (60 and 49%
respectively). Also, for the 1:1 ratio reaction, a larger amount of the
ethylated axial alcohol (80%) and reduced axial alcohol (17%) were
observed although the amount of ethylated and reduced alcohols remained
the same (25 and 75% respectively) as those observed for the Et 2A1X
reactions. A major difference in the ratio of products occurred for the
reaction in which the reagent to ketone ratio was 3:1. In this case, 91%
of the reacted ketone was converted to the reduced alcohol with 32% of that
a result which was greater than twice that observed for the Et 2A1I reac-
tion. Hence EtAlI2 is the most stereoselective; however, with such a low
yield of alcohols, it is an impractical reagent.
Phenylation of Ketones
The phenylanalogs of the above reagents were prepared according
to the procedure described in the Experimental Section. They were all
allowed to react with ketone (I) in benzene at different mole ratios
with the results listed in Table 56. Because the above reactions showed
no enhanced stereoselectivity in TBF, only benzene reactions were
carried out. The 3:1 reagent to ketone ratio reactions showed no en-
hanced stereoselectivity just as for the methylation or ethylation
reactions described above (92-95% equatorial alcohol). The most interesting
reactions were those where the reagent to ketone ratio was 1:1. As the group
size increased from Cl to Br to I, the amount of recovered ketone (50 to 55
to 65%, respectively) and equatorial alcohol (60 to 63 to 70%, respectively)
also increased. Also, PhAlI 2 , when allowed to react with ketone (I), con
tinued the trend with 80% recovered ketone but the percent of phenylated
equatorial alcohol increased to 75% compared to 25% for the axial alcohol.
254
This once again points out ,the fact that as the reagent becomes larger
the stereoselectivity also increased but the yield is so low that the
reagent is impractical.
CHAPTER IV
CONCLUSION
The major conclusion from this study is that the greater the steric
bulk of the RnAIX3-n
compounds the greater the stereoselectivity, with the
best reagent being R2A1I. But, because of low yields, the use of substituted
organoaluminum reagents is impractical for the addition to enones and/or
ketones.
255
Table 48. Reactions of MenAlIC3_n Compounds With Enone
REAGENT REAGENT
SOLVENT TEMP (QC)
TIME (hr)
RECOVERED, ENONE'(%) u
ALIMATION PRODUCTS TOTAL 1,4 1;2 ENONE
Me3A1 1 THE RT 24 20 75 3 97
3 20 76 1 99
1 45 8 -17 78 3 97
3 16 77 2 98
1 Benzene RT 24 2 97 5 95
3 1 96 1 99
1 45 8 2 94 4 96
3 2 93 1 99
Me2 AlC1 1 THE 24 25 71 6 94
3 24 70 5 95
1 Benzene 21 70 7 93
3 20 70 6 94
Table 48 (Continued)
Me 2A1Br 1
3
I
THE
Benzene
RT 24 35
33
32
58
58
57
20
20
20
3 31 58 21
Me 2A11 1 THE RT 24 53 42 98
50 42 98
1 Benzene RT 24 47 47 99
46 47 99
M_eAlI2 THE RT 24 61 32 100
60 32 100 1 Benzene RT 24 55 39 100
3 54 39 100
79
0
0
REAGENT TEMP TIME RECOVERED REAGENT ENONE. ALKYLATION PRODUCTS (%) a
E SOLVENT ( ° C) (hr) ENONE'(%) b TOTAL '1;4 1,2
80
80
80
0
o 3 o
0
0
0
0
0
0
REAGENT REAGENT
SOLVENT T FM' ( °C).
Table 48 (Continued)
TZ RECOVMED, (hr) ENONE (%) u ENO NE
H 2Al (NPr i) 2 1 THE RT 24 8 9
87
Benzene RT 24 76
3 75
tie2 AlOCH3 1 THF . RT 24 75
3 75
Benzene RT 24 70
65
Me2 i10Bu t THE RT 24 91
91
Benzene RT 24 72
70
ALKYLAT ION PRODUCTS (7. ) a TOTAL 1, 4 1,2
Table 48 (Continued)
ALKYLATION PRODUCTS (%) 4 REAGENT ENONE SOLVENT (QC) (hr) ENONE'(%) b TOTAL - 1;4 1,2
REAGENT TEMP TIME RECOVERED
THE RT 24 91
91
89
89
3
0 0 0
I
3
Benzene
0
a) Normalized as per cent 1,4-product + per cent 1,2-product = 100%.
b) Yield determined by glc and based on internal standard.
Table 49. Reactions of Me2A1T With Other Enone6 in Benzene and THE at Room Temperature for 24 Hours in a 2:1 Ratio.
RECOVEREDa METHYLATION PRODUCTS (%) b ENONE' -SOLVENT ENONE (%) ' " TOTAL' ' ''' 1;2'"'
260
THE
Benzene
39 40 55 45
20 39 61 39
THE
PK-N....N)?•flt Benzene 33 60 67 33
15 75 68 32
THE
Benzene
50 40 65 25
45 • 38 66 24
a) Yield determined by glc and based on an internal standard.
b) Normalized as % 1,4-product + % 1,2-product = 100%.
Ph3P
Ph3P
RECOVERED METHYLATION PRODUCTS Mb SOLVENT ENONE.(%)a TOTAL 1,4 1,2
THE 0 0 0
Benzene 0
THE 0
Benzene 4 0 0
THE 0 32 0 100
Benzene 0 47 0 100
THE 91 0 0
Benzene 28 64 92
THE 90 2 0 100
Benzene 89 15 85
THE 90 0
Benzene 89 0 0
•n-Bu3 P
HMPA
•n-Bu3P
HMPA
COORDINATING ENONE AGENT
• Table 50. Reactions of Me AlWith Enone (I) an alone (II) in the Presence of Coordinating Agents at Room Temperature for 24 Hours.
261
a) Yield determined by glc and based on an internal standard.
b) Normalized as % 1,4-product + % 1,2-product = 100%.
Table 51. Reactions
REAGENT
of EtriAlK.3-n Compounds With Enone (II).
ADDITION. PRODUCTS (%) TEMP TIME RECOVERED
SOLVENT • (°C) (lit) ENONE (%)C TOTALa ' 1,4It 1 . 2b
REDUCTION PRODUCTS (%)
. .
TOTALa . 1,4b, 1,2b REAGENT ENONE
1 THE RT 24 0 93 Trace 100 1 0 100 Et3A1
3 0 92 Trace 100 1 0 100 1 45 8 0 90 Trace 100 1 0 100 3 0 90 Trace 100 1 0 100
1 Benzene RT 24 0 92 Trace 100 1 0 100 3 0 93 Trace 100 • 1 0 100 1 45 8 0 90 Trace 100 1 0 100 3 0 91 Trace 100 1 0 100
Et 2A1C1 1 THF RT 24 0 90 . Trace 100 3 0 100
3 0 90 Trace 100 2 0 100
1 Benzene RT 24 0 85 1 99 4 0 100
3 0 84 1 99 3 0 100
Et2A1Br 1 THF RT 24 0 85 1 99 5 0 100
3 0 84 1 99 5 0 100
1 Benzene RT 24 0 78 3 97 6 0 100
3 0 75 4 96 6 0 100
REAGENT SOLVENT
TEMP ( °C)
TIME (hr)
RECOVERED ENONE (%)
Table 51 (Continued)
ADDITION PRODUCTS (%)
TOTAL '1;4 . 1 .2
REDUCTION PRODUCTS (%)
b b TOTAL !,4 144 1 2 REAGENT ENONE
Et2A1I 1 THE RT 24 0 80 15. 85 12 0 100
3 0 79 17 83 13 0 100
1 Benzene RT 24 0 75 18 82 13 100•
3 0 73 19 81 14 100
EtA1I2 1 THE RT 24 0 50 27 73 30 100
49 28 72 31 0 100
1 Benzene RT . 24 Trace 45 28 72 35 100
3 Trace 46. 29 71 3'6 100":
Et2A1NPr 2 1 THE RT 24 0 35 15 85 10 0 100
3 0 34 16 84 11 0 100
1 Benzene RT 24 0 31 16 84 11 0 100
3 0 31 15 85 11 0 100
Et2A1
3
1
3
Table 51 (Continued)
ADDITION PRODUCTS (%) REDUCTION PRODUCTS (7.) REAGENT TEMP TIME RECOVERED
REAGENT ENONE SOLVENT (°C) (hr) . ENONE . (%) c TOTALa 1;4b
TIT RT 24 0 30 13
0 29 13
Benzene RT 24 0 25 14
0 25 15
'1;2b 'TOTALa - 1;4b 1,2
87 100
87 7 0 100,,
86 0 100
85 0 100
a) Normalized as % addition pr6ducts + reduction products = 100%.
b) Normalized as% 1,4-product + % 1,2-product = 100%.
c) Normalized as % total products + % enone = 100%.
Table 52. Reactions of Ph 1AIK3..n Compounds With Enone (II).
REAGENT REAGENT
SOLVENT TEMP
(°C) TIME (hr)
RECOVERED "ENONE'(%)a
ADDITION PRODUCTS TOTAL 1,4
(%)b
1,2 ENONE
Ph3Al 1 THE RT 24 0 80 60 40
3 0 79 60 40
45 8 0 79 61 39
3 0 80 61 39
1 Benzene RT 24 0 80 60 40
3 0 81 59 41
45 8 0 79 59 41
78 59 41
Ph2AlC1 1 THE RT 24 0 78 64 36
3 0 78, 65 35
I Benzene. RT 24 0 77 65 35
3 0 75 65 35
REAGENT SOLVENT TEMP (°C)
Table 52 (Continued)
TIME RECOVERED — (hr) ' ENONE'.(%)a
ADDITION PRODUCTS (%) TOTAL : 1,4 , 1,2 c
REAGENT 'ENONE
Ph2A1Br 1 THE RT 24 0 75 88 12
3 0 75 89 11
I Benzene RT 24 0 74 89 11
3 0 73 90 10_
Ph2AlI 1 THE RT 24 0 64 100
66 100
1 Benzene RT 24 0 66 100
65 100
PhAlI2 THE RT 24 0 45 100
44 100
1 Benzene RT 24 0 40 100
39 100 0
REAGENT 'REAGENT
SOLVENT TEMP °C)
Table 52
TIME (hr
(Continued)
RECOVERED ENONE (% )a
ADDITION PRODUCTS (%)b TOTAL 1 4 L 2 'ENONE'
Ph AlNPr 2 2 THE RT 24 70 25 74 26
3 68 25 74 26
Benzene RT 24 65 27 75 25
3 66 26 74 26
Ph2AlOCH3 THE RT 24 70 25 75 25
72 26 74 26
I Benzene RT 24 65 25 75 25
3 65 25 75 25
Ph2 A10But THE RT 24 75 22 76 24
73 22 76 24
Benzene RT 24 75 22 76 24
75 22 76 24
Ph2A1( I THE
3
RT
Benzene
Table 52 (Continued)
REAGENT TEMP TIME RECOVERED REAGENT " . ENONE SOLVENT (°C (ht •ENONE(%)a
24 90
90
24 89
90
a) Yield determined by glc and based on an internal standard.
b) Normalized as % 1,4-product + % 1,2-product = 100%.
ADDITION PRODUCTS (%)b
'TOTAL 1 4 '1 2
11 81 19
11 81 19
10 80 20
11 80 20
Table 53. Reactions
REAGENT
of Me n AIX
REAGENT
3-n Compounds With 4-t-Butylcyclohexanone, Ketone (I).
ADDITION PRODUCTS (%) b TEMP TIME RECOVERED AXIAL EQUATORIAL
SOLVENT (°C) ,(hr) KETONE (%) 'TOTAL ALCOHOL ALCOHOL KETONE
Me3A1 1 THE RT 24 50 47 85 15
3 50 48 86 14
45 8 45 45 85 15
3 40 50 83 17
1 Benzene RT 24 50 45 76 24
3 1 97 11 89
1 45 33 59 75 25
3 0 95 9 91
Me 2A1C 1 1 THE RT 24 80 17 85 15
3 80 1.8 83 17
1 Benzene RT 24 70 26 57 43
3 17 79 10 90
Table 53 (Continued)
REAGENT ' REAGENT .
SOLVENT TEM? (°C)
TIME (hr)
RECOVERED. KETONE ' (%)"
ADDITION PRODUCTS (% ) b AXIAL . EQUATORIAL
TOTAL ' ALCOHOL ALCOHOL KETONE
Me 211 Br 1 THE RT 24 83 15 80 20
3 83 16 80 20
1 Benzene RT 24 80 12 40 60
3 25 66 8 92
Me2Al I I THE RT 24 93 7 79 21
3 90 7 79 21
1 Benzene RT 24 87 6 20 80
3 40 53 7 93
MeAl 12 1 THE RT 24 97 80 20
3 97 1 80 20
Benzene. RT 24 91 4 18 72
3 50 46 6 94
T REAGENT
SOLVENT TEMP
. (°C)
Table 53
TIME (1.1t)
(Continued)
RECOVERED. KETONE . (%)7
ADDITION PRODUCTS (7)b
. AXIAL EQUATORIAL
TOTAL ALCOHOL ALCOHOL KETONE
Me 2A1NPr2 , 1 THE RT 24 60 20 85 15
3 55 20 85 15
1 Benzene RT 24 55 1 100 0
3 47 1 100
Me 2AlOCH3 1 THE RT 24 80 5 85 15
3 70 10 86 14
1 Benzene RT 24 75 5 86 14
3 68 11 85 15
m'a2 Al0But THE RT 24 61 5 80 • 20
3 59 5 80 20
1 Benzene RT 24 60 5 85 15
58 8 86 14
REAGENT KETONE
1
Me Al( 100
100
100
0
Table 53 (Continued)
ADDITION PRODUCTS (%) b 'REAGENT TEMP TIME RECOVERED AXIAL EQUATORIAL
(%)a ' "TOTAL ALCOHOL : ALCOHOL ' SOLVENT ((C) ) KETONE
THE RT 24 55
50
Benz ene RT 24 50
40
THE RT 24
Benzene RT 24
1 100
0
1 100
4 100
a) Yield was determined by glc and based on an internal standard.
b) Normalized as % axial alcohol + % equatorial alcohol = 100%.
Table 54. Reactions of Ile Al With Ketone (I) in the Presence of Coordinating Agents at Roan Temperature for 24 Hours in a 1t1:1 Ratio.
COORDINATING AGENT SOLVENT
-RECOVERED -a KETONE (%)"''''TOTAL. AXIAL' EQUATORIAL
Ph3 P Benzene 70 3 75 25
THF 75 5 85 15
n-Bu P Benzene 70 5 75 25
THF 70 5 75 25
UMPA Benzene 60 5 75 25
DHE Benzene 55 4 75 25
a) Yield was determined by glc and based on an internal standard.
b) Normalized as % axial alcohol + % equatorial alcohol = 100%.
273
Table 55.
REAGENT
Reactions of Et nAlK3_n
REAGENT. .
Compounds With 4-t7Bdtylcyclohexanone, Ketone (I).
ADDITION PRODUCTS (%) REDUCTION PRODUCTS (%) TEMP TIME RECOVERED . EQUATORIAL AXIAL -EQUATORIAL CC) (hr) KETONE (%) C - TOTAL°. - ALCOHOLb ALCOHOLb -TOTAL4 ALCOH0b- - ALCOHOLb KETONE SOLVENT
Et3A1 1 THE RT 24 35 65 88 12 35 22 78
3 0 80 88 12 20 22 78
1 Benzene RT 24 46 64 79 21 36 19 81
3 0 72 8 92 28 29 71
Et2A1C1 Benzene RT 24 31 31 65 35 69 20 80
3 6 41 27 73 59 26 74
Et 2 AlBr 1 Benzene RT 24 30 29 70 30 71 20 80
3 6 40 20 80 60 25
Et 2A1I 1 Benzene RT 24 30 27 75 25 73 20 80
3 5 43 15 85 57 35 65
EtA1I2 1 Benzene RT 24 60 25 80 20 75 17 83
3 49 9 35 65 91 32 68
a) Normalized as % alkylation alcohols + % reduction alcohols = 100%. b) Normalized as % axial alcohol + % equatorial alcohol = 100%.
c) Normalized as % total alcohol + % ketone = 100%.
Table 56. Reactions
'REAGENT
of PhriAIX3 _a
REAGENT
Compounds With 4-t-Butylcyclohemanone,- Ketone (I).
ADDITION PRODUCTS (%) b TEMP - TIME RECOVERED AXIAL EQUATORIAL
:SOLVENT (olo (hr) KETONE (%) a ' TOTAL ALCOHOL ALCOHOL KETONE
Ph3Al 1 Benzene RT 24 45 50 51 49
0 95 8 92
Ph2A1C1 1 50 43 40 60
3 J nn7%.1
(IC 7.J
Ph2A1Br 1 55 : 40 45 55
3 15 80 6 94
Ph2A1I 1 65 30 30 7CY
3 25 65 5 95
PhAlI2 80 15 25 75
3 40 50 5 95
a) Yield was determined by glc and based on an internal standard.
b) Normalized as % axial alcohol + % equatorial alcohol = 100%.
LITERATURE CITED
L. H. C. Brown and H. M. Hess, J. En v Chem., , 2206(1969).
2. E. C. Ashby and B. Cooke, J. A.M. Chen. Soc., 90, 1625(1968).
3. S. Masamune, P. A. Rossy and G. S. Bates, J. Am. Chem. Soc., 6452(1973).
4. J. E. McMurry and M. P. Fleming, J. Am. Chen. Soc., 96, 4708(1974).
5. E. C. Ashby, J. J. Lin and R. Kovar, J. Elm. Chem., 41, 1939(1976).
6. E. C. Ashby and R. Kovar, Inorg. Chen., 16, 1437(1977).
7. G. H. Posner, "Organic Reactions", Vol. 19, p. 1 (1972).
8. a) H. C. Brown and G. W. Kabalka, J. Au. Chem. Soc., 92, 712(1970). b) H. C. Brown and G. W. Kabalka, J. MIL Chem. Soc.,- 92, 714(1970).
9. G. W. Kabalka and R. F. Daley, J. An. Chem. Soc., 95, 4428(1973).
10. E. C. Ashby and G. Heinsohn, J. Org. Chem., 39, 3297(1974).
11. A. E. Jeffery, A. Meisters and T. Mole, J. Organometal. Chen., 101, 345(1974).
12. E. C. Ashby and R. D. Schwartz, J. Chem. Educ., 51, 65(1974).
13. D. F. Shriver, "The Manipulation of Air-Sensitive Compounds", McGraw-Hill, New York, N. Y., 1969.
14. T. Mole, Australian J. Chem., 16, 794(1963).
15. D. L. Schmidt and E. E. Flagg, Inorg. Chem., 6, 1262(1967).
16. E. C. Ashby and S. A. Noding, Inorg, Chem., (in press).
17. E. C. Ashby, J. J. Lin and J. J. Watkins, J. Org. Chem., 42, 1099 (1977).
18. W. Kirmse, J. Knist and H. J. Ratajczak, Chem. Ber. 109, 2296(1976).
19. J. Ficini and A. Maujean, Bull. Soc. Chim. Fr., 219(1971).
20. E. C. Ashby, J. Laemmle and P. V. Roling, J. Org. Chem., 38, 2526 (1973).
276
277
21. a) O. L. Chapman and R. W. King, J. An. Chem. Soc., 86, 1256(1964). b) R. J. Ouellette, J. Am. Chen. Toc., 86, 4378(1964). c) G. D. Meakins, R. K. Percy, E. E. Richards, and R. N. Young, J. Chem. Soc., C, 1106(1968). d) J. Battioni, W. Chodkiewicz and P. Cadiot, C. R. Acad. Sci., Ser. C, 264, 991(1967). e) J. Battioni, M. Chapman and W. Chodkiewicz, Bull. Soc. Chim. Fr., 976(1969). f) J. Battioni and W. Chodkiewicz, Bull. Soc. Chin. Fr., 981(1969). g) J. Battioni and W. Chodkiewicz, Bull. Soc. Chin. Fr., 1824(1971).
22. J. R. Luderer, J. E. Woodall and J. L. Pyle, J. fla. Chen. 36, 2909 (1971).
23. a) E. C. Ashby and S. Yu, J. Chem. Soc., D, 351(1971). b) E. C. Ashby, S. Yu and P. V. Roling, J. Org„':Chem., 37, 1918(1972).
24. a) , E. C. Ashby and S. Yu, J. Chem.'Soc., D, 351(1971). b) E. C. Ashby, S. Yu. and P. V. Roling, J. Org. Chem., 37, 1920(1972).
25. J. L. Namy, C. R. Acad. Sci., Ser. C, 272, 1334(1971).
26. J. B. Melpolder and R. F. Heck, J. Orz .. Chem., 41, 165(1976).
VITA
Stephen Alfred Noding was born on October 5, 1947, in Slayton,
Minnesota and subsequently attended public school in District L268,
Des Moines River Township, Murray County, and in Slayton, Minnesota.
graduated from Slayton High School in June 1965 and attended Augustana
College in Sioux Falls, South Dakota, from 1965 to 1970 when he received
the Bachelor of Arts Degree in Chemistry and Mathematics. After two
years' military service in the United States Army, he entered the School
of Chemistry, Georgia Institute of Technology in September 1973 to pursue
a Ph. D. under the direction of Dr. Eugene C. Ashby.
The author has accepted a position of employment with Dow Chemical
USA, Louisiana Division at Plaquemine, Louisiana.
278